Working Papers in Economic History
February 2018 WP 18-03
Getting to Denmark’: the Role of Elites for Development
Peter Sandholt Jensen Markus Lampe Paul Sharp Christian Volmar Skovsgaard
Abstract We explore the role of elites for development and in particular for the spread of cooperative creameries in Denmark in the 1880s, which was a major factor behind that country’s rapid economic catch-up. We demonstrate empirically that the location of early proto-modern dairies, so-called hollænderier, introduced onto traditional landed estates as part of the Holstein System of agriculture by landowning elites from the Duchies of Schleswig and Holstein in the eighteenth century, can explain the location of cooperative creameries in 1890, more than a century later, after controlling for other relevant determinants. We interpret this as evidence that areas close to estates which adopted the Holstein System witnessed a gradual spread of modern ideas from the estates to the peasantry. Moreover, we identify a causal relationship by utilizing the nature of the spread of the Holstein System around Denmark, and the distance to the first estate to introduce it, Sofiendal. These results are supported by evidence from a wealth of contemporary sources and are robust to a variety of alternative specifications..
Keywords: Institutions, technology, knowledge spillovers, landowning elites, cooperatives, Denmark JEL Classification: N53, O13, Q13
Markus Lampe (corresponding author): Institute for Economic and Social History, Vienna University of Economics and Business (Welthandelsplatz 1, Bg D4, 1020 Vienna, Austria) and Instituto Figuerola. E-mail: [email protected] https://www.wu.ac.at/geschichte/institut/personal/lampe-markus/ Publisher: Carlos III University of Madrid. Figuerola Institute of Social Sciences History www.uc3m.es/if
UNIVERSIDAD CARLOS III DE MADRID • c/ Madrid 126 • 28903 Getafe (Spain)• Tel: (34) 91 624 97 97 Site: http://portal.uc3m.es/portal/page/portal/instituto_figuerola/home/publications/working_papers
Series: Working Papers in Economic History ISSN: 2341-2542
Electronic version of these working paper series available on: http://hdl.handle.net/10016/16
This work is licensed under a Creative Commons Attribution-NonCommercial- NoDerivatives 4.0 International License. ‘Getting to Denmark’: the Role of Elites for Development
Peter Sandholt Jensen, University of Southern Denmark
Markus Lampe, Vienna University of Economics and Business
Paul Sharp, University of Southern Denmark, CAGE, CEPR
Christian Volmar Skovsgaard, University of Southern Denmark
Abstract: We explore the role of elites for development and in particular for the spread of cooperative creameries in Denmark in the 1880s, which was a major factor behind that country’s rapid economic catch-up. We demonstrate empirically that the location of early proto-modern dairies, so-called hollænderier, introduced onto traditional landed estates as part of the Holstein System of agriculture by landowning elites from the Duchies of Schleswig and Holstein in the eighteenth century, can explain the location of cooperative creameries in 1890, more than a century later, after controlling for other relevant determinants. We interpret this as evidence that areas close to estates which adopted the Holstein System witnessed a gradual spread of modern ideas from the estates to the peasantry. Moreover, we identify a causal relationship by utilizing the nature of the spread of the Holstein System around Denmark, and the distance to the first estate to introduce it, Sofiendal. These results are supported by evidence from a wealth of contemporary sources and are robust to a variety of alternative specifications.
Keywords: Institutions, technology, knowledge spillovers, landowning elites, cooperatives, Denmark
JEL codes: N53, O13, Q13
We would like to thank Philipp Ager, Sascha Becker, Per Boje, David de la Croix, Carl-Johan Dalgaard, Jeremiah Dittmar, Alice Fabre, Giovanni Federico, James Fenske, Nicola Gennaioli, Ingrid Henriksen, Anton Howes, Pablo Martinelli, Joel Mokyr, Nathan Nunn, Cecilia Garcia Penalosa, Jared Rubin, and James Simpson for helpful comments and suggestions. Moreover, we would like to thank conference and workshop participants at the 2017 World Congress of Cliometrics in Strasbourg, the Agricliometrics II in Zaragoza, the CAGE/CEPR Economic History Conference in Abu Dhabi, the 2016 Social Science History Association meeting in Chicago, the CREA 2016 workshop in Luxembourg, the MEHR end of semester workshop 2016 at Copenhagen University, Simposio de la Asociación Española de Economía 2015 in Girona, the 2017 Economic History Society Conference, the 2017 workshop on ‘Elite human capital and the role of modernity: The East versus the West’ in Marseille, as well as seminar participants at the University of Santiago, the Catholic University of Chile, the Central Bank of Chile, Lund University, George Mason University, Remnin University, the Central University of Finance and Economics (Beijing), the University of South Australia, the London School of Economics, the University of Duisburg-Essen, the University of Bonn, and the University of Siena. Finally, we would also like to thank Dorte Kook Lyngholm from the Dansk Center for Herregårdsforskning (Danish Research Center for Manorial Studies) for sending us some of the data on the estates. Markus Lampe acknowledges funding from Fundación Ramón Areces, and Paul Sharp gratefully acknowledges funding from the Danish Council for Independent Research: this paper is part of his Sapere Aude grant no. DFF-6109-00123.
1
1. Introduction
A substantial literature in economics examines the impact of elites on the growth trajectories of societies across time and space through various channels. While the impact on economic and political institutions is well- established, the role of elites for knowledge diffusion is not as well understood. In this paper, we shed new light on how elites may foster growth through this channel. We focus on a specific example centered on the Kingdom of Denmark, and the emergence of a modern dairy industry based on a new technology, the automatic cream separator (a steam-powered centrifuge), and a new institution, the cooperative creamery, which propelled the country towards prosperity in the last decades of the nineteenth century (for a brief account, see Henriksen 1993). After the foundation of the first cooperative creamery in 1882, hundreds of cooperative creameries spread throughout the whole country within a decade. Massive increases in productivity followed, production boomed, Denmark captured a large share of the important UK market for butter and other agricultural products, and witnessed rapid economic catch up with the leading economies of the day, as traditional suppliers of agricultural goods such as Ireland and the Netherlands lost market share. An important point motivating the present work is the rapidity with which the cooperatives spread, with the first wave over by 1890; see figure 1.
In a light-hearted manner, Francis Fukuyama (2011) has described the issue facing developing countries as the problem of ‘getting to Denmark’, a metaphor for a society characterized by wealth, the rule of law, good governance, and related virtues. But how did Denmark get to Denmark? For economic historians, the rise of the cooperatives is usually considered to be a turning point, but this is questioned in a recent book on the subject (Lampe and Sharp 2018), where we argued that it was rather the end result of a long period of agricultural enlightenment, as a modern, scientific form of agriculture spread into and throughout the country, propelled initially by a group of elites1 on landed estates one hundred years earlier. The intermediate steps are described in detail in the book, and are summarized in the present work, which constitutes a reduced form test of their hypothesis. Lampe and Sharp’s (2018) account is a significant revision of our understanding of the Danish development path, with important policy implications, largely due to the role Denmark often plays as a poster boy for the idea that countries can develop through cooperation in the countryside, as is perhaps best exemplified by the classic account commissioned by the Food and Agricultural Organization of the United Nations after the Second World War (Skrubbeltrang 1953). However, the idea that a country with a large number of peasants and cows (for example Ireland at the turn of the twentieth century, or India after the Second World War) could simply cooperate its way out of underdevelopment has been severely criticized by Lampe and Sharp (2018) for exactly
1 Or at least their administrators, tenants and/or skilled dairy workers, the latter mostly women, who traditionally dominated that sector.
2 the reason presented here, that successful cooperatives stood on the shoulders of at least a century of accumulated improvements in agricultural theory, practice, and teaching.
Figure 1: Location of Cooperative Creameries in 1890
Source: Own work, based on Bjørn (1988).
Beyond the policy implications, it is certainly useful to understand what allowed Denmark to grow rapidly from some point in the second half of the nineteenth century2, eventually to become one of the richest countries in the world, with many other desirable socioeconomic characteristics (low inequality, high levels of ‘happiness’, etc. – see Lampe and Sharp 2018). The discussion among Danish historians has largely focused on whether agriculture can be seen as the leading sector in this process, creating demand and growth in other sectors, such as industry (supplying the centrifuges and refrigerators used in dairying for example) and services3. Revisionists argue that
2 Reliable GDP/capita data for Denmark do not unfortunately exist for this period. See however the work on real wages by Khaustova and Sharp 2016. 3 In particular transportation, where the Danish shipping giant Mærsk is a direct descendent of the firms that carried the agricultural exports to the UK.
3
the developments in agriculture were simply part of a wider process of development which touched all sectors of the economy (see for example Boje 2016).
Whatever the case, the success of Danish agriculture was extraordinary. Within a few years at the end of the nineteenth century, Denmark captured a significant share of the important UK market for animal foodstuffs. By around 1900, over ninety percent of Denmark’s total production of butter and pork went to the UK. This was also significant from the UK side. Danish exports made up an increasingly large share of total imports, increasing from around 15 percent in the 1870s to over 43 percent before the First World War for butter, from under 1 percent to almost 50 percent over the same period for bacon, and from under four percent to almost one fourth for eggs (Henriksen 1993, p. 156). This Danish success is usually set within the context of the American ‘grain invasion’ from the 1870s (O’Rourke 1997), when cheap exports of American grain flooded Europe, promoting a backlash of protectionism. Denmark, like the UK, chose to remain open, however, and successfully exploited its comparative advantages, using the cheap grain as fodder for increased animal production, and outcompeting traditional suppliers of the UK market, such as the Irish.4
Danish agricultural products (Lurpak butter and Danish bacon) still today constitute important components of the English breakfast, although Denmark is no longer so dependent on the UK market, and Danish agricultural products can be found all over the world. Denmark’s current status as an ‘agricultural superpower’5, dominated by massive firms such as Arla (a dairy cooperative) and Danish Crown (a food, especially meat, processing firm previously also a cooperative until 2010), is usually traced back to the aforementioned developments in the 1880s. As we will discuss in more detail below, at this point a new technology, the steam-powered automatic cream separator made it possible to use milk which had been transported over long distances to be processed in a central production facility, and the voluntary associations of Danish peasants, the cooperatives, sprang up to take advantage of this possibility. Thus, modern Denmark emerged based on a democratic, cooperative countryside, providing something of a role model to other agricultural countries around the world.
As attractive as this narrative might appear, it begs the obvious question as to why other countries could not so easily follow similar strategies. As noted above, it also remains to be explained how hundreds of butter factories could spring up in a few years in the 1880s, and how dominance in agricultural exports could be so rapidly consolidated. The existing literature (basically Henriksen 1999, inspired by Ó Gráda 1977) has attributed the rapid
4 Thus, however, as Boserup (1992, p. 57) noted long ago, this development was crucially dependent on Britain’s decision to remain a free trader, which allowed this process to happen in a way which is perhaps inconceivable for developing countries today, which are constrained by the protectionist policies of the US and the EU in particular. 5 Economist, January 4, 2014 ‘Bringing home the bacon: Tiny Denmark is an agricultural superpower’.
4
diffusion of the cooperative dairy movement in Denmark mostly to pre-existing cow densities. In other country- commodity specific studies, the scale of production prior to the introduction of cooperatives has also been highlighted, apart from other product-specific factors and access to transportation networks. Recent internationally comparative studies (Fernández 2014) have highlighted the importance of social capital (or trust) proxied by a variety of variables, especially (low land) inequality and (protestant) religion. This follows important work by O’Rourke (2006, 2007), who argued that it was the absence of conflicts and the egalitarianism of the Danish population which distinguished it from Ireland, where cooperation emerged later and less successfully. This homogeneity can in turn be traced back to more secure property rights in Denmark, due to the centralization of government and land reforms in the eighteenth century, which made it more difficult for the elites to exploit the peasantry, who also enjoyed more secure assets, with the consequence that violent conflict was less likely, and networks and social capital could form more easily. While religion and social fractionalization have proven to be important in other countries, this can arguably not explain the adoption pattern within Denmark given the extremely homogenous population.
Lampe and Sharp (2018) argue, in contrast to the traditional accounts, that the cream separators and cooperatives marked the end of the story of the modernization of Danish dairying, where the former was the final piece of the jigsaw, allowing the latter to form – an unintended end-result of something which happened over a century earlier, as elites moved into Denmark. This can be traced to developments following Denmark’s rather numerous military defeats at the hands of Sweden6, which led to a considerable loss of territory, and financial difficulties for the Crown and the privatization of almost all the crown estates by 1740 in order to raise revenue. A rather inconclusive debate about how to take advantage of this situation in order to introduce reforms and modernize agriculture ensued, but with the ultimate result that many were sold to enterprising Danish and German nobles, merchants, and others. A well-functioning market for land meant that this process also spread to other estates, beyond the former crown estates. Many were bought up by elites from Schleswig and Holstein, who saw opportunities for vastly improving the productivity of the Danish countryside. They brought with them the relatively sophisticated agricultural system they knew from home, the Holstein System, which importantly for the emergence of the cooperatives introduced the idea of centralizing dairy production in a central facility. This in turn set in motion an early ‘enlightened’ approach to agriculture, including modern standards of bookkeeping and accounting (important for science and for discovering what is most profitable to produce); the establishment of knowledge institutions: apprenticeships, agricultural schooling, extension, etc.; and experimentation both on the estates as well as at specialist research institutions – in fact the first centrifuge in Denmark was trialed on one of
6 In fact, Denmark and Sweden hold the record for having fought the largest number of wars.
5
the estates. Importantly, however, the estate owners of the 1880s did not support the cooperatives directly, seeing them as competition, and even actively campaigned against them. Nevertheless, Lampe and Sharp (2018) argue that it was a trickle down of ideas from the elites in the 1700s that allowed the cooperatives to emerge so successfully a century later. In our reduced form of their hypothesis, we exploit that not all parts of the country were equally influenced by the elites, which in turn explains the specific spatial pattern of the cooperatives.
We thus demonstrate econometrically that the pattern of adoption of cooperative creameries in Denmark followed the introduction of proto-modern dairies by agricultural elites on estate farms. In the Duchies of Schleswig and Holstein, ruled by the King of Denmark in personal union until 1864 when they were lost to Prussia, an intensified crop rotation system with an important dairy component was developed on the large manorial estates known as Koppelwirtschaft in German, or kobbelbrug in Danish. It became the dominant field system in the Duchies in the 1700s, and included unprecedentedly large herds of milch cows and the invention of an innovative new centralized system of butter production, the hollænderi, with unparalleled standards of hygiene and equipment (Porskrog Rasmussen 2010a). These innovations – collectively known as the above mentioned ‘Holstein System’ when the crop-rotation was combined with the dairy unit – came relatively late to Denmark, but when they did they gradually transformed Danish agriculture. We show how the specific organizational and technological innovation of estate dairy production came to Denmark in the 1760s and quickly increased in importance, although this adoption was unequal across the country, and led to an uneven spread of emulation by common farmers in the following decades, a common pattern for the diffusion of innovations in early modern societies, as highlighted by Mokyr (2009) for the British ‘agricultural enlightenment’. Hence, the main alternative use of the same resources, cattle-fattening, within the production system of Koppelwirtschaft, was still discussed and seen as viable in the run-up to the spread of dairy cooperatives. Subsequent scientific debate led to further advances, including accurate bookkeeping, better breeds of cows, and better feed. Agricultural societies, schools, and journals were established. More generally, it became firmly established that Denmark’s comparative advantage lay in dairying, and butter production in particular. As we will discuss, and summarize below, we have previously (Lampe and Sharp 2018) established a narrative account – based on substantial use of primary archival and other sources – of how this innovation spread and trickled down to the peasantry through general and specific education, research, diffusion of examples of best practice and the establishment of channels of marketing.
Our econometric analysis relies on a novel database for which we have collected a substantial amount of data from a combination of primary and secondary sources. For the most part we have this data on the grid level, and we divide Denmark into 38,370 1x1km grid cells. Controlling for a large number of other relevant determinants of cooperation, including indicators of land quality and suitability for dairying, we find that each grid cell’s exposure
6 to cooperative creameries in 1890 (measured as the sum of the inverse distance to every cooperative from that location) correlates with that cell’s treatment by the elites, measured by the sum of the (tax assessed land quality adjusted) sizes of all estates with hollænderier (and thus the Holstein System) in 1782 weighted by their distance to the grid cell. Then, we demonstrate causality using the fact that the Holstein System spread throughout the country in a particular way, starting with the first, which was established on an estate called Sofiendal, the distance to which we use as an instrument. The use of distance as an instrument follows studies on the spread of ideas such as Dittmar (2011) on the diffusion of the printing press and Becker and Woessmann (2009) and Akçomak et al. (2016) on the spread of religious practices that favor literacy. Our results are not just statistically but also highly economically significant, implying that an increase of one standard deviation in the elite influence increases the cooperative exposure 42 percent of the mean exposure in one of our preferred specifications. They are also robust to a variety of alternative specifications, including using data on the parish level and using the estates (with vs. without the Holstein System) as the unit of analysis.
In contrast to much other similar research, the intermediate steps during the century between the elites establishing the hollænderier by 1782 and the end of the initial spread of the cooperatives in 1890 are not a black box, and we discuss them briefly in this paper, although a full account is provided by Lampe and Sharp (2018). Certainly, however, an important prerequisite to the rapid establishment of cooperative creameries was through an increasing interest in dairying by the peasantry, which we are able to measure thanks to occasional agricultural censuses taken during the nineteenth century. Thus, we are also able to demonstrate that one channel through which the elite influence trickled down to the peasantry was via increased cattle densities in areas which had been treated more by the elites. Thus we conclude that the early and rapid spread of the cooperative creameries, which in turn are often considered to have propelled Denmark toward modernity, and were an important precursor of the country’s agricultural success today, was due to the accumulation and spread of knowledge over a century. Developments on the Danish estates are in this way put in their rightful place as the starting point of the Danish agricultural revolution, which was to change Denmark forever.
Besides the literature on the role of elites for development, on agricultural cooperation, and on the development of Denmark, the present work is also closely connected to recent studies that show the long-run impact of the adoption of agriculture (Olsson and Hibbs 2005; Putterman 2008; Comin et al. 2010, Cook 2014a) and major productivity improving implements like the (heavy mouldboard) plough (Andersen et al. 2016), as well as complementing the emerging literature on the effects of new crops on productivity, population and economic
7
growth, and political stability (e.g. Nunn and Qian 2011, Cook 2014a, Cook 2014b, Dall et al. 2014, Chen and Kung 2016, Jia 2014, Bustos et al. 2016).7
The following section provides a brief review of the relevant literature on the role of elites for development. Section 3 describes the Holstein System as it emerged in Schleswig and Holstein, and considers its spread into and throughout Denmark. Section 4 provides an empirical analysis of the impact of this for the emergence of the cooperatives more than a century later. Section 5 tests the robustness of the results, and finally Section 6 concludes.
2. A brief literature review on the role of elites for development
A long running theme in growth and development is how elites influence the long run economic trajectories of societies (Amsden, DiCaprio and Robinson 2012). It is undeniable that elites exert disproportionate influence on how institutions are designed and factor endowments are used, but it is less well known how their actions feed into the development process. For the case of agricultural elites, the existing literature has provided conflicting views on the role of large landowners in economic development. On the one hand, the concentration of agricultural resources in the hands of large landowners and accompanying high levels of land inequality are often seen as an impediment to development. Engerman and Sokoloff (2002) stress the interaction between factor endowments and the resulting impact of land inequality and how higher land inequality leads to agricultural elites who favor slavery and extractive institutions, which produces poor economic outcomes, see also Acemoglu, Johnson and Robinson (2001) ad Bannerje and Iyer (2005). Galor, Moav and Vollrath (2009), Baten and Juif (2014) and Cinnirella and Hornung (2016) demonstrate that high land inequality causes elites to block investment in human capital. Others have stressed that high land inequality limits the scope for agricultural cooperation both through lower social capital (Fernández 2014) and through direct crowding-out (Henriksen 1999). On the other hand, in contexts in which property rights are poorly defined, large landowners can ‘shelter’ dependent peasants from extractive state institutions (Dell 2010) and effectively lobby for better provision of collective goods and infrastructure than politically weak peasant communities (Dell 2010, Dell 2012).8
7 In addition, our work connects to literatures such as the role of immigration for technology and knowledge transfer, the significance of local knowledge spillovers from large to small firms, and to ‘new new’ trade theory, which suggests that firms that export are more productive, and that more productive firms export more. 8 In Denmark, during the nineteenth century, the property rights of peasant farmers became increasingly more secure through enclosure and formation of inheritable property. Hence we do not believe – and find no evidence – that estates per se should have had a necessary facilitating function for the establishment of cooperatives. On the other hand, while large landowners in Denmark might have been interested in blocking education and other rights for peasants between the 1780s and the 1880s, as evidenced by the short-lived second
8
By contrast, we stress that agricultural elites may spread knowledge, which then subsequently aids development in the agricultural sector. In other words, our work suggests that agricultural elites may also be knowledge elites, who facilitate later development. Recent work by Squicciarini and Voigtländer (2016) demonstrates that knowledge elites played a significant role in the industrialization of France by e.g. running businesses themselves or exchanging knowledge with entrepreneurs. Our work emphasizes the importance of knowledge spill-overs and agricultural enlightenment (Mokyr 2009, ch. 9), and shares some similarities with Hornung’s (2014) work on high- skilled immigration of Huguenots into Prussia. He shows that this led to higher productivity in the textile sector and interprets this as evidence of an effect of diffusion of technology. We focus on agricultural elites and their impact on the part of the agricultural sector that led to an economy-wide take-off.
3. The evolution of modern dairying and its spread to Denmark
In this section, we provide the argument that we set out to test. We first explore the origin of the Holstein System. We next describe how it spread into and across Denmark. As will become clear in Section 4, we will apply our knowledge of the spread across Denmark to identify the causal effect of early elites on the spread of cooperative creameries. We finally discuss how the spread of proto-modern dairying eventually propelled the spread of cooperatives.
The origin of the Holstein System and Koppelwirtschaft
The Holstein System and Koppelwirtschaft more generally was a ‘collective invention’ by estate owners and their administrators in sixteenth-century Holstein and Schleswig, the German-speaking part of the Danish monarchy, in order to overcome the fundamental problem of intensified organic agriculture, i.e. how to sustain production and yields in the long run by obtaining sufficient fertilizer from animal husbandry. This challenge was met in different parts of Europe in different ways, especially in modern-day Belgium, Northern France, the Netherlands and Britain, where ‘convertible husbandry’ systems developed (Mokyr 2009, p. 173; Jones 2016). The Holstein model consisted of changing the traditional three-field rotation with outlying pasture areas into an eleven-field rotation, thus alternating the use of individual fields between pasture and grain cultivation over eleven years. 9 This way, extensively used grazing areas (pastures) were included in the crop rotation by changing the traditional design of fields and the crop rotation itself in a way that allowed for sustained grain yields and sufficient fodder for the
serfdom in the late eighteenth century, effectively, the centralization and professionalization of government and the Constitution of 1849 reduced the scope for such action. However, since long-term effects of elite blockage might have persistent outcomes, we control for the share of land under the direct control of estates (demesnes).
9 For more details on alternative systems see Lampe and Sharp (2018) and the references provided there.
9
animals, normally in the form of summer pasture and winter hay – all this at the same time as production surpluses were exported from rural areas in order to sustain growing urban populations. This system was relatively more focused on animal production than alternative systems, in part because soils were particularly suited for fertile grasslands in Holstein and Schleswig, which in the sixteenth-century had focused on oxen fattening and horse breeding. In part, the evolution of Koppelwirtschaft in the seventeenth and eighteenth centuries seems to be the reaction of estate owners to an improvement in the relative prices of dairy products versus grain and oxen (Porskrog Rasmussen 2010a, p. 180), which led to intensified collaboration with available specialized immigrants from the Netherlands and their descendants so as to develop a strong dairy sector (Porskrog Rasmussen 2003, p. 447). Since these specialists originated from Holland, the tenants involved in dairying became known as hollænder (and their dairies as hollænderier), even if they were not of Dutch descent.10
Koppelwirtschaft was introduced in the demesne farming of large manorial estates, and not in peasant agriculture, because these were the most commercially oriented agricultural units, the most likely to be able to sustain the considerable capital investments and labor efforts (via corvée or hired labor) required for reorganizing the fields, and also the ones with the largest freedom to act under the institutional framework of the time. Many manorial estates were managed by relatively professional staff (Porskrog Rasmussen 2010a, p. 182), and dairying became a professionalized subset of demesne farming activities, in which specialist hollænder managed specialized dairy equipment under a regime of well-specified dairy lease contracts (Drejer 1925-33, p. 181-2; Iversen 1992, p. 76- 77; Porskrog Rasmussen 1987, pp. 63-65 and Lampe and Sharp 2018). Hollændere were normally a couple, with the woman in charge of dairy production and the man supervising the feeding of the cows and the transportation of raw materials and produce to and from the dairy (Hansen 2006). Moreover, the Holstein system implied a proto-modern dairy with a centralized production facility for separation of cream from milk and production of butter much like the cooperative creameries a century later. 11 Under the Holstein System, many estates in Holstein and Schleswig came to have very large herds of several hundred cows, even exceeding herd sizes in Holland (Porskrog Rasmussen 2010a, pp. 181-2).
Within Holstein, Schleswig, and subsequently Denmark,12 the Holstein system thus led to a model of production with advances on the estates in different parts of agriculture and thus the opening of a gap in quality between
10 Bieleman (1996) gives an account of the sophisticated dairy sector in the Low Countries during the Dutch ‘Golden Age’. 11 The hollænderi would also have practical independent rooms, a strong focus on hygiene, cows milked at particular times (and milked dry), control of the temperature of the cream, so it could be skimmed and churned at the right time, and care would be exerted at all times from milking to packaging. 12 Koppelwirtschaft and hollænderier also spread to the east to Mecklenburg in the eighteenth century, but relatively little research is available on this, apart from Schröder-Lembke (1978, 65-67), who portrays the troubled figure of Mecklenburg Koppelwirtschaft pioneer Joachim Friedrich von der Lühe and his Panzow estate. See also Jones (2016, pp. 95-97).
10
estate and peasant producers emerged in the extent to which produce was regularly marketed (Lampe and Sharp 2015a) and in terms of grain yields, butter quality, etc. (Bjørn 1988, p. 159; see also Lampe and Sharp 2014, 2015b). In the Kingdom of Denmark itself, from the Middle Ages until the seventeenth century, estates as well as peasant farms typically only had as many cows as they needed to feed the household, and more sophisticated dairy products were imported from Holland (Appel and Bredkjær 1924-33, pp. 279-80).13 Even for the latter half of the 1700s, scholars often highlight the low proportion of cows (and bulls) relative to horses in the use of pasture in Denmark. Hertel (1920, 149-51) for example, estimates the cattle to horse ratio at only 1.4:1 in the 1770s14, much less than the 4:1 in 1914.
The introduction of the Holstein System to Denmark
An important prerequisite to the introduction of the Holstein System was the redistribution of land throughout the eighteenth century. In the 1600s Denmark consisted of a large number of Crown Estates, under the direct administration of the monarch, smaller estates owned by the nobility, as well as many medium sized subordinate farms belonging to estates (Porskrog Rasmussen 2003, p. 8). From the 1600s and into the 1700s, the bad finances of the crown, largely as a result of continuous wars against Sweden until 1721, meant that monarchs were forced to sell off more and more land, until by the 1740s almost all the crown estates were privatized (Frandsen 2005, p. 58, 74-76), with a final touch of privatizations in around 1770. At the same time, the introduction of absolutism in Denmark in 1660 had weakened the privileges of the traditional nobility and opened up the possibility of estate ownership to non-nobles (Lampe and Sharp 2018, Linvald 1912). This situation encouraged discussions, also in the government, on how to introduce reforms and a general modernization of agriculture (Jensen 1998, p. 37-8; Feldbæk 1988, p. 19).
The introduction of Koppelwirtschaft was to be the result of this debate, although most saw it simply as a means to increase grain yields. This is best illustrated by a famous quote by Adam Gottlob Moltke, effectively prime minister from 1746 to 1766 and generally credited with introducing the Holstein System to Denmark (Jensen 1998, p. 92), from a plan devised in 1746 for King Frederik V: ‘Agriculture in these lands seems to be still very backward. I keep myself assured that, if the soil here would be worked as is custom in other countries, especially in Holstein, the land could yield twice as much as it has produced hitherto.’ (quoted from Porskrog Rasmussen 2010b, p. 9 and note 1).. In the context of his ascent to Lord Chamberlain for Frederik V in 1746, Moltke also received the
13 The word hollænderi entered the Danish language apparently from the eighteenth century with the spread of Koppelwirtschaft into Denmark (Drejer 1925-33, p. 138). Some prior examples of larger dairy units, managed by hollænder from Holstein or the Netherlands have been mentioned in the literature, but with little sustained impact (Drejer 1925-33, pp. 140-143; Skrubbeltrang 1978, p. 120; Frandsen 2005, pp. 46-47, 146). 14 When, however, cattle pests had done much to reduce cattle stocks.
11
large estate of Bregentved in Southern Zealand, and up to 1751 bought four more nearby estates: Turebyholm, Juellinge, Tryggevælden and Aslev (Porskrog Rasmussen 2010b, p. 11). 15 In 1759, he came, under fortuitous circumstances, to own the estate of Niendorf near Lübeck in Holstein, on which the Holstein System was firmly established. He sold it two years later with a large profit (Porskrog Rasmussen 2010b, p. 19-21) and took the former leaseholder of Niendorf, Johann Matthias Völckers, to his estates on Zealand to become his administrator and agricultural reorganizer there. Völckers started on the newly established farm of Stenkelstrup (later named Sofiendal after Moltke’s second wife) to implement an exact copy of Holstein Koppelwirtschaft with the layout of the eleven fields, the original crop rotation and a hollænderi, and finished this in 1766. He then continued to reform Moltke’s estates of Alslev, Turebyholm and the Bregentved main estate up to 1767 and Juellinge in the early 1770s. Most of Moltke’s reorganized estates were then, as before, leased in auctions to interested leaseholders, including Völckers himself (Jensen 1998, p. 49-51, Porskrog Rasmussen 2010b). In reports he wrote for the king in the 1780s to highlight his role as a reformer, Moltke claimed that the value of his lease contracts in 1787 had increased by more than 200 percent since the introduction of the Holstein System in comparison to the 1740s, although modern research has qualified this somewhat since estate leases had generally increased over the period (Porskrog Rasmussen 2010b, p. 26-7). There is, however, no doubt that Moltke’s reorganization increased the capitalized value of his estates, and his descendants are in fact still the largest noble landowners in Denmark16, according to a list published by the Danish public broadcaster, DR, in relation to a recent debate on lowering inheritance taxes for family-owned businesses.
Moltke was imitated by his neighbors. For example, the Løvenborg estate was reorganized in 1767 with Völckers as expert, and the Gisselfeld estate, adjacent to Bregentved, in 1768 (Porskorg Rasmussen 2010b, 27; Jensen 1998, 52). In 1769 the estate of the Vemmetofte Jomfruekloster was reorganized, with Völckers as consultant to its administrator (Linvald 1905-08, p. 250; Prange 1971, p. 552). Gradually Moltke’s example was followed in other parts of Denmark, and by 1800 most demesnes were using Koppelwirtschaft (although not necessarily with the dairy unit), while peasant agriculture still largely relied on the medieval three-field system (Falbe Hansen 1889, p. 10; Bjørn 1988a, p. 35; Frandsen 2005, p. 90).17 In fact, the list of estates having a hollænderi in 1782 reads like
15 During the next decades, Moltke would own estates in all parts of Denmark as well as in Schleswig and Holstein and become the largest landowner in the Monarchy (see the map in Porskrog Rasmussen 2010b, p. 14). His cultivation reforms in Denmark were centered mostly on the aforementioned estates on Zealand and the ones he bought between 1763-5 on Funen. 16 The net worth of the Moltke family (still based in Bregentved) was almost 1 billion Danish kroner (around 150 million US dollars) in April 2017. DR, ‘Grafik: De største adelige godsejere i Danmark’, https://www.dr.dk/nyheder/penge/grafik-de-stoerste-adelige- godsejere-i-danmark, retrieved December 8, 2017. 17 In the late 1760s, he and Völckers also developed a version of Koppelwirtschaft for the villages dependent on his estates which respected traditional common land rights (fællesskab) (Porskorg Rasmussen 2010b, 30-35). It did, however, not spread as fast and widely as its estate demesne counterpart.
12
something of a who’s who of nobility: old, new, high, low, Danish and foreign, and includes some of the leading reformers of the time.18
Over the subsequent decades, despite a general economic and agricultural crisis as a consequence of the Napoleonic Wars and state bankruptcy, Koppelwirtschaft continued to spread across Denmark. Lampe and Sharp (2018) provide a detailed account of this, based on a large number of primary and secondary sources, principally for the first half of the nineteenth century a series of books by Begtrup (1803, 1806, 1808) and a series of reports on the state of agriculture in each county commissioned by the Royal Agricultural Society and published between 1826 and 1844 (Dalgas 1826), the latter of which reveal for example that by the late 1830s or 1840s good dairy practices were spreading to the peasantry in the south of the island of Zealand close to where Moltke first introduced the Holstein system in 1766, and that estates further west adopted the Holstein System more slowly, thus supporting our narrative that both the adoption of the Holstein System by estates, and the early spread of good practices to the peasantry, was more frequent closer to Sofiendal.
In fact, by the 1840s, the transformation of Denmark was already well underway, decades before the first cooperative. Thus, the prominent German travel writer Johann Georg Kohl observed that the Holstein System had spread throughout the country, even to the opposite end of the country in Northern Jutland, where he noted that many farms had switched from oxen-raising to dairying. He was impressed by the scientific nature of this progress, and noted that important articles on dairying from the Duchies were reprinted all over Denmark. In conclusion, he stated his belief that Denmark would eventually converge on the Duchies, and that they would finally integrate completely with the Kingdom to become a land ‘not of milk and honey, but of milk and butter’ (Kohl 1846, pp. 58- 60). Other foreigners, even from the UK, marveled at the scale of the operations, such as the British writer, Samuel Laing, who wondered at the ‘regularity, arrangement, cleanliness and the vast scale of all the operations [which] give the impression rather of a great manufactory of butter and cheese than of a farm’ (Laing 1852, p. 124).
The spread of knowledge and cooperative creameries
18 Apart from Moltke and his son Joachim Godske Moltke with nine different estates, there was also for example the famous reformer Christian Ditlev Frederik Reventlow, as well as Christian von Benzon, Christine Sophie Friis and her husband Erhard Wedel-Friis, and Frederik Siegfried Christiansen Rantzau, all with three estates; as well as many others with one or two estates, including Theodor Adeler, Sigfred Victor Raben-Levetzau, Lorentz Christian Ernst Cederfeld de Simonsen, Jørgen Wichmand Wichfeld, Hans Henrik von Eickstedt, Godske Hans von Krogh, Frederik Ludvig Christian Beenfeldt, Christian Ahlefeldt-Laurvig, and members of the Rosenkrantz, Gyldencrone, Hardenberg, Holstein, Lüttichau, Stampe, Raben (the family of Moltke’s wife), Juel and Finneke(-Blixen) families etc. However, among them we also find parish priests (Niels Frederiksen Amager, whose widow owned Gedsergård), merchants (Hans Bergeshagen Hincheldey of Valnæsgård), apparently a pharmacist (Henrik Schmidt of Haraldskær), and the son of an estate tenant who started as an estate tenant himself (Jens Lange of Løjtved).
13
In the next section, we demonstrate empirically that the estate creameries had a trickle-down effect on the peasantry, consistent with the historical narrative. Lampe and Sharp (2018)19 provide a detailed account of the chain of events and the developments which connect the two end points of the reduced form analysis presented in this paper: the establishment of the elite-owned estate creameries in the late eighteenth century, and the emergence of the peasant-owned cooperative creameries from 1882. Thus, we describe in detail the developments which were initiated by the elites since the establishment of the Holstein System, and document the links that emerged between the estates and the peasantry. Here, we provide a concise, self-contained summary.
A program of agricultural reforms went alongside the spread of the Holstein System in the late eighteenth century, with the end result that for example serfdom (or ‘adscription’) was abolished and land enclosures were put in place, firmly establishing private property rights in the countryside. The completion of these reforms by the first years of the nineteenth century coincided however with the Napoleonic Wars, which were particularly devastating for Denmark. Copenhagen was almost completely destroyed in a British bombardment in 1807, and the Danish fleet was captured, and in the terms of the peace Denmark lost Norway to Sweden in 1814. The Danish state went bankrupt, and a profound period of uncertainty followed, although a second wave of elites (again, often from Holstein and Schleswig) began to establish themselves in Denmark from the 1820s, and built on the reforms of the eighteenth century. Of particular importance was the early introduction of an ‘enlightened’ approach to dairying and agriculture more generally, involving accurate measuring and recordkeeping, combined with sophisticated bookkeeping and accounting. This allowed first of all for a scientific and experimental approach to agriculture, answering questions such as how best to feed, breed, and milk cows. Second, accounting allowed for a better idea of profitability, allowing for profit-maximizing behavior and for example the discovery that specialization in dairying was the best strategy. Lively debates on methodology and the implications of the findings made played out in the Danish agricultural press in particular from the mid-nineteenth century. We argue that the knowledge built up in this process laid the foundation for the rapid spread of smallholder dairying later on, in particular because – apart from the example of specialization – it was increasingly taught through specialized agricultural schooling and apprenticeship programs.
The Danish estates owners were also joined from the second half of the nineteenth century by a new set of elites, merchants, who established the marketing channels necessary for taking advantage of markets abroad, especially in industrializing Britain, and encouraged quality improvements to obtain more marketable produce for export.
19 In part based on previously published articles: Henriksen, Lampe and Sharp (2011, 2012), Lampe and Sharp (2014, 2015a)
14
This was in part by encouraging medium-sized farms and smallholders outside the realm of the estates to centralize production, initially through a model of privately owned so-called community creameries in the 1860s (Bjørn 1977, McLaughlin and Sharp 2015). Little information remain on these, however, since they were small- scale and suffered from the problem that peasant producers owned just a few cows, and their milk production could not easily be transported to a central production facility. This was solved, however, with the invention of the automatic cream separator in the late 1870s20, which was able to separate the cream from transported milk using centrifugal force. The cream separator thus finally allowed peasants to enjoy the benefits of centralized production and marketing pioneered by the hollænderier more than a century before, this time largely in the form of cooperative creameries. They did not enjoy an easy start, however. The first coops in southwestern Jutland met with great skepticism from the agricultural establishment, that is, estate owners. Thus, the chairman of the dairy committee of the United Jutland Agricultural Associations (and member of the board of the Royal Agricultural Society of Denmark) commissioned an instructor from the agricultural college of Ladelundgaard to travel around eighteen of them in order to demonstrate their inferiority compared to the privately-owned community creameries which he had previously reported on (Petersen 1885; Henriksen 1999). Although his report reached the opposite conclusion to that which its commissioners had hoped for, there can be little doubt that the estates themselves were not promoting the cooperative form as such.
4. Persistency and the spread of the cooperatives
To test for the influence of the elites on the location of cooperative creameries a century later, we examine whether areas closer to estates using the Holstein System were more likely to have cooperative creameries nearby after the first wave of cooperatives ended around 1890. Specifically, we divide Denmark into 38,370 1 x 1 km grid cells to be able to pick up the very local geographical variation in the location of cooperative creameries and the estates of the elites and be able to account for potential cofounding factors at a very detailed level.21 In practice, we calculate measures of the degree to which a cell is near hollænderier and cooperative creameries.
Our main proxy for the spillovers from elites to peasants is the variable elites 1782, constructed in the spirit of Harris’ (1954) ‘market potential’ (mp) measure as the sum of the tax assessed land quality adjusted sizes of all estates with hollænderier weighted by their distance. We calculate this mp for all grid cells. The reasoning behind
20 In fact, it seems that the principle that cream could be separated using centrifugal force was discovered in Germany in 1864, but the crucial refinements were made in the Duchy of Holstein in 1876. Separators based on this design were then launched by rival Danish and Swedish firms in 1878/9 (Pedersen 1999, p. 51). 21 Population and barley suitability are the only control variables that we cannot directly generate at the grid level and hence we attribute the parish level population counts to each cell within the parish and with FAO’s more aggregated barley suitability rasters. We assume the suitability to be constant within FAO’s 5 arc minute resolution which corresponds to approximately 9 x 5 km raster cells in Denmark.
15
this strategy is that the influence of the elites increases with the estate size, but decreases with distance. In other words, the closer and larger the estates, the more influence they would have had in terms of knowledge spillovers. When assessing the relationship in this way, however, one needs to bear in mind that any surviving hollænderier on estates would have been competitors to the first cooperatives (see Henriksen 1999). 22 The estates with hollænderier are given in figure 3, together with all other estates in 1782. To measure the extent to which a grid cell is exposed to cooperative creameries, we also construct a market potential measure for dairies in a similar fashion and construct the variable cooperative creamery exposure as the sum of cooperative creameries weighted by the inverse distance from the grid cell to the cooperative.
Table 1 provides a brief description of the variables included in the regressions below, as well as summary statistics and sources. Apart from cooperative creamery exposure and elites 1782, we also construct several other variables, which might also contribute to explaining the location of cooperatives in 1890. The first is distance to the first cooperative creamery as the cooperatives might conceivably have spread from there. Second we control for the share of a grid cell that is occupied by estate demesnes (with or without hollænderier) in the late eighteenth century. Since this land was directly cultivated by estates, it is much less likely that (peasant) cooperatives would be founded there. We also control for the proximity of all estates to capture the general political and social effects of the presence of large landowners on peasant cooperation, which might be negative or positive, following the various results in the literature discussed above. We separately control for (pre-privatization) crown ownership of estates as these estates were typically larger and more centrally located and hence might have had an independent and perhaps stronger impact on the surrounding peasant society. Next we add controls for ideas in the half century leading up to the spread of the cooperative creameries in the form of proximity to a number of late innovators and to folk high schools. The former identifies estate owners who participated in a commission set up by the Danish Royal Agricultural Society in 1858 and were particularly influential. The latter controls for the impact of the folk high schools which traditionally have be seen as a main driver of the cooperative movement due to their important role in educating peasants23 and as they shared many of the same ideas and philosophies as those behind the cooperative movement.
Figure 3: Location of hollænderier among all estates in 1782
22 Unfortunately, data on which estate creameries survived until 1882/1890 are unavailable. As noted above, the agricultural establishment was initially hostile to the peasant cooperatives and as such surviving creameries work against our hypothesis making the results presented below conservative estimates. 23 The courses supplied by the folk high schools were decided by the head of each school but typically the peasants learned about hygiene in the production of milk, cultivation of plants and more general knowledge about democracy and how to participate in society.
16
¨
Source: Own work based on Andersen (1963)24, Christensen (1886) and Roholt (2012). One might further argue that the Holstein System was simply established in areas with previous dairy know-how and cow herds. We thus introduce a number of controls to take account of this possibility, including the presence of a grass field system in the seventeenth century, historical butter production, clover cultivation in 1805 and barley suitability25, the latter of which captures the main alternate use of the land.26 First, Frandsen (1983) gives direct information on dairying, that is, the amount of in-kind rent payments in butter made by peasants in 1662 per unit of land.27 Although this says little about demesne production of butter on estates, it might be positively
24 Andersen identifies the presence of Hollænderier based on 652 estate accountings (“amtsmandsindberetninger”) from the Ministry of Finance (“Rentekammeret”), 1782. We match the estates with Hollænderier with the 1770 list of estates from the Danish Research Center for Manorial Studies, supplemented by the list from Christensen (1886, appendix B). The latter is a transcription of a manuscript without title, date or signature preserved in the Royal Library in Copenhagen (Univ. Mskr. Add. 230). Christensen dates it to around 1770. For more on this see the discussion by Linvald (1912, p. 150, fn. 1). 25 Strictly speaking present day potential yields of rain-fed barley, from FAOs GAEZ database (2002). As shown by Andersen et al. (2016) present day potential barley yields correlates strongly with the level of barley tenant payments under the feudal system in 1662. As the historical data do not provide full coverage we use the measure of present day potential yields. All results are robust to using the subsample for which historical data are available. 26 It might also potentially capture its availability as fodder. Under Koppelwirtschaft dairying and grain production can be considered to be complements, but the cooperatives also imported grain and concentrates from overseas. 27 This information was collected by the government for several commodity payments in order to construct a proxy for land productivity as a basis of immediate taxation before the actual land survey was carried out.
17
related to the later establishment of cooperatives if there is persistency in dairying patterns among the peasantry (which hollænderier might just have taken advantage of). Second, Frandsen (1983) reports information on the prevalence of field-grass-systems in agriculture in the 1680s at the time of the large land quality (and use) assessment for a new land-based taxation system – this might be positively related to dairying, but it might more likely proxy for the use of cattle-raising. Third, as clover was an important part of Koppelwirtschaft and the whole Holstein system, we control for the share cultivated with clover in 1805 as a further control for pre-existing conditions for dairying. Finally, we add distance to the coast as a simple measure for openness in the form of market access given that the cooperatives were heavily export oriented.
To further control for market access we add distance to Copenhagen, parish level population density, proximity to market towns, distance to the rail road in 1890, and distance to the Ox Road, where the latter controls for proximity to the main export route for cattle which was the main alternative use of similar resources.
<< Table 1 around here >>
As a first step to disentangle whether the knowledge of the elites spread to the peasantry over time Table 2 provides baseline OLS results from estimating the following regression equation:
= + 1782 + + . (1)
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑖𝑖 𝛼𝛼 𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽 𝑖𝑖 𝑿𝑿′𝒊𝒊𝜹𝜹 𝜖𝜖𝑖𝑖 where is a grid cell, is cooperative creamery exposure, 1782 is our proxy for the influence𝑖𝑖 of the elite𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 on cell as ex𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐plained𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐above, is a vector of control variables𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 described above, is the associated vector of coefficients𝑖𝑖 and is the error ter𝑿𝑿𝒊𝒊 m. The standard errors are corrected for clustering𝜹𝜹 at the parish level. We also compute and report𝜖𝜖𝑖𝑖 Conley standard errors to account for potential spatial autocorrelation not captured with dependence within the parish. In most specifications, contains region fixed effects.
𝑿𝑿𝒊𝒊
<< Table 2 around here >>
In all specifications in Table 2, the coefficient on 1782 is positive and statistically significant. This implies a
positive effect on the likelihood that cooperative𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒creameries were established in proximity to hollænderier established by elites. The introduction of regional fixed effects in column 2 has very little impact on the estimated
18
coefficient. 28 In column 3, we control for the distance to the first cooperative creamery, and find that the coefficient is largely unchanged as compared to columns 1 and 2. As controls for estates without hollænderier and the presence of an estate demesne in the grid cell are included in column 4, the size of the effect is more than cut in half but is still highly significant. Controlling for crown estates, late innovators and the presence of folk high schools has little impact on the coefficient on 1782 in column 5. The same is true when we include barley
suitability, historical butter production, the share𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 of𝑒𝑒𝑒𝑒 the area growing clover, historical presence of the grass field system, and the distance to the coast in the set of control variables in column 6. Finally, when we add control variables for market access (Distance to Copenhagen, Population density, presence of a market town, Distance to the railway and Distance to the Ox road) in column 7, the estimated coefficient on 1782 is reduced substantially, but remains strongly significant. The estimated coefficient in column 7 of Table𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 2 implies that an increase of one standard deviation in the elite influence increases the likelihood for a grid cell to be exposed to cooperative creameries by (0.00243*70,81)=0.17, or 9 percent of a standard deviation in the cooperative creamery exposure using the most conservative estimate. This indicates that the effect of the elites is not only statistically, but also economically, significant.
The results also produce some interesting findings for the control variables. Being closer to the first cooperative creamery is positively associated with the emergence of cooperatives as one might expect. The presence of estates or crown estates nearby is also positively associated with the emergence of cooperatives but not if the location is on an estate demesne. Further, the presence of folk high schools and late innovators are also positively associated with the presence of cooperative creameries as expected. All market access and geographical control variables enter with the expected signs except for distance to the coast and grass-field-system. Cooperatives were more intensely established further away from the coast and in areas without a grass-field-system in the 1680s. One potential explanation for the former is that the distance to the sea had become less important with the arrival of the railroad, which itself could have been an important locational determinant for many cooperatives. The result on grass-field-system may be interpreted as a legacy of the historical optimal use of land. In the seventeenth century those parts of the country that were most fertile were cultivated using the traditional three-field-system whereas the less fertile sandy soils especially found in western Jutland were primarily used for grazing. With the advent of the cooperative movement, the fertile parts of the country could support more cows and in turn more cooperative creameries.
28 The regional fixed effects are for 21 historical counties when we use the grid level data. For the much smaller parish and estate samples, we use fixed effects for 5 larger historical regions (Jutland, Funen, Zealand, Lolland-Falster and Bornholm). These fixed effects capture, among other things, that some regions historically were subject to serfdom longer than others.
19
Even though we include an extensive set of control variables, the concern that omitted variables determine both the location of hollænderier and cooperatives could remain. To address this, we propose an instrumental variable identification strategy, where we instrument 1782 by the distance to Moltke’s estate, Sofiendal, where the
Holstein System was first established. This is 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒consistent𝑒𝑒𝑒𝑒 with our story that the system spread through Denmark inspired by Moltke (and his administrator, Völckers) as well as the historical literature cited above. Hence, we estimate the following instrumental variables model:
= + 1782 + + . (2)
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑖𝑖 𝛼𝛼 𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽𝛽 𝑖𝑖 𝑿𝑿′𝒊𝒊𝜹𝜹 𝜖𝜖𝑖𝑖 1782 = + + + . (3) ′ 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑖𝑖 𝜃𝜃 𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝛾𝑒𝑒𝑛𝑛𝑛𝑛𝑛𝑛𝑙𝑙𝑖𝑖 𝑿𝑿 𝒊𝒊𝚪𝚪 𝜇𝜇𝑖𝑖 where we include the great circle distance29 to Sofiendal in our first stage (3) as our excluded instrument in (2).
In Table 3 we present the results of the first stage. The table follows the same structure as Table 2 and we control for the same variables as above. We notice that the coefficient on the distance to Sofiendal is always negative and strongly significant. Moreover, as shown at the bottom of Table 4, the F-test of instrument relevance is always well above 10, as per the usual rule of thumb. Thus, the instrument is highly relevant and there is no sign that distant to Sofiendal is a weak instrument. In column 4 in Table 3, It is seen that hollænderier, among other things, were established in places nearer estates in general but away from the estates of the late innovators. Proximity to crown estates does not influence the location of hollænderier. Higher suitability for barley production increases the likelihood of being close to hollænderier as do places further away from the coast and closer to the historical Ox Road as seen in columns 6 and 7. Importantly, the inclusion of the control variables does not change the statistical significance of the relation between the location of the hollænderier and Sofiendal.
<< Table 3 around here >>
We present the instrumental variable (or two-stage-least-square) estimates of equation (2) in Table 4. Compared to Table 2, the coefficient on the 1782 variable is now larger and more stable as control variables are added.
Thus, when we rely on the plausibly,𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒exogenous component of 1782, we obtain large and significant effects that can be interpreted as the causal impact of the landed elites𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 and𝑒𝑒𝑒𝑒 the associated spread of knowledge on the emergence of the cooperative creamery movement. To a large extent the coefficients on the control variables are
29 We have estimated all models using cost distance instead of great circle distance. Results are very similar in terms of significance as well as magnitude.
20
like those reported in Table 2. It is, however, worth noticing that when using the instrumental variable strategy 1782 remain significant while other estates now have a negative impact and crown estates are only
30 borderline𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 significant.
A potential threat to identification is the validity of the exclusion restriction of the instrument. Thus, even if we have included many control variables, there may still be remaining concerns about whether this restriction is violated. In order to further investigate the robustness of our results to this threat we use the ‘plausibly exogenous’ technique of Conley et al. (2012). In Figure C1 in the appendix, we apply this technique to gauge how large a potential direct effect of the instrument (Distance to Sofiendal) needs to be to render the IV estimate on 1782 insignificant. The estimates suggest that some omitted variable that is also captured by Distance to
𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒Sofiendal𝑒𝑒𝑒𝑒 needs to explain about 86 percent of the overall reduced form effect of the distance to Sofiendal to render the 2SLS estimate on 1782 insignificant. As this is very high, we conclude that while it is possible that the direct effect is of this order𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 of𝑒𝑒𝑒𝑒 magnitude, this does not seem plausible (for further details see Appendix C).
<< Table 4 around here >>
To evaluate the importance of the elite based explanation relative to other potential explanations we calculate standardized coefficients in Table 5. When compared to all the potential cofounders, we see that 1782 has much larger explanatory power. In fact, the beta coefficient on 1782 explains more than all𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒other the 16 covariates combined. In column 7, a one standard deviation increase𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 in the 1782 variable leads to a 1.73 standard deviation increase in the cooperative creamery exposure variable, or𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 1.73*2.00𝑒𝑒𝑒𝑒 = 3.46 which is 42% of the mean exposure. This is our preferred estimate as the instrumental variables estimate arguably measures the causal effect.
<< Table 5 around here >>
If the presence of hollænderier had a persistent effect for a century before the first cooperatives, we would expect that this meant a gradual spread of the ideas used on the estates to the wider peasant population, due to the
30 We have also estimated models in which we control for calorie adjusted crop yields as constructed by Galor and Özak (2016) for the post 1500 period. This measure enters with a negative coefficient and has little impact on the estimated effect of 1782.
21 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
traditional links between the estates and the surrounding peasantry, and reflected by our reading of the contemporary literature. We can quantify this by considering the increase in the number of milch cows around the country in the intervening period. In 1760 there were 270,000 milch cows in Denmark, increasing to 335,000 in 1774, and 450,000 in 1810 (Drejer 1962, p. 22, Jensen 1998). Moreover, in 1837 we have parish level data from the first (surviving) animal census, which puts the total level at 578,000 in 1837. In 1861, there were 756,834 milch cows in the animal census. By 1881, the year before the first cooperative creamery was founded, there were 898,790. If we are to believe the persistency story, the local density of cows should have remained fairly constant before 1882. In fact, the correlation coefficients between the densities in 1837 and 1861 and 1881 are all around 0.9. It then remains to demonstrate that the location of the hollænderier also explains the pattern we observe in the cow densities. To do this, we employ the same empirical strategy as above, but with the cow densities in 1837 at the parish level as the outcome variable. Again, the relationship is very strong and robust – a greater influence from the elites implies greater cow densities (see Table 6). This result is consistent with the work of Henriksen (1999), who shows that the share of cows supplying a cooperative correlates with cow density, which implies that the peasantry was also turning to dairying in areas close to hollænderier, imitating the elites as the narrative above implies.
<< Table 6 around here >>
Looking across the table, we note that adding control variables does not change the significance of 1782.
Importantly, columns 6 and 7 show that the relationship between cow densities and 1782 is not𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 driven𝑒𝑒𝑒𝑒 by past specialization in butter production by the peasantry. 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
5. Robustness
In this section, we investigate the robustness of our results in four ways. First, we show that the results are similar for the location of cooperatives in 1914. Second, we use alternative units of observation and find that the results also hold at the parish and estate levels. Third, we test the relevance of our instrumental variable at the estate level and finally, we investigate whether our main results hold water in a model without the inclusion of fixed effects for regions.
We start by testing the robustness of our results in the temporal dimension. Hitherto we have focused on the time around the ending of the first wave of the cooperative creamery movement around 1890. One might argue that
22
the results are sensitive to this specific year and therefore we select another point in time to test the hypothesis. Specifically, we select 1914, just before the First World War changed the landscape. Tables 7 and 8 present the results from this robustness test. Focusing on the standardized coefficients in Table 8 we see they are statistically significant, though slightly smaller than those for 1890, again emphasizing the persistence of the impact of the eighteenth century elites.
<< Tables 7 and 8 around here >>
Next, we show that the relationship between the elites and the cooperative movement is present also at the parish level. We use the parish level to demonstrate that results also hold when historical and larger units are used. Moreover, the parish level corresponds to the historical level at which population data were collected and hence we do not need to disaggregate these data. We use the same strategy as presented above but use parishes as the unit of observation instead of grid cells.31 Table 9 shows the second stage results. We see that the results are robust to using parishes as the unit of observation. Reassuringly, the results are very similar to the grid level results in terms of both magnitude and significance.
<< Table 9 around here >>
Next we consider estates as an alternative unit of observation. We show that the relationship between the elites and the cooperative movement is also present if we apply a similar strategy at the estate level. Using the estate as the unit of observations confers the advantage that we now compare estates with estates which are likely to be more similar units than e.g. parishes. In this way, we reduce heterogeneity between the units of observations, but of course end up with fewer observations. We estimate the model using a similar outcome variable as before, but the distance to the cooperative creameries now being measured from each estate. We code a dummy for whether an estate used the Holstein System in 1782 and construct similar control variables as in the setups described above.32 Thus, the variation comes from the plausibly exogenous component in the establishment of a
31 Parish level descriptive statistics can be found in Appendix B. 32 Estate level descriptive statistics can be found in Appendix A.
23
hollænderi on an estate. Table 10 shows the second stage results. We see a positive and significant relation between the presence of a hollænderi on an estate and the proximity of cooperative creameries. In column 7, the significance level is only at the five percent level, whereas in other columns, significance is at the one percent level. Moreover, the F statistic is just above 5 in column 7 and thus below the conventional rule of thumb. However, as noted by Cameron and Trivedi (2005), an F statistic above 5 is also sometimes applied as a rule of thumb .
<< Table 10 around here >>
We found distance Sofiendal to be a relevant instrument at the grid and parish levels above, so we think the weaker relation is likely due to a lower number of observations in column 7 of Table 10. To test the relevance or the predictive power of distance to Sofiendal in explaining the geographical distribution of hollænderier across estates we randomly allocate the 224 hollænderier to the 791 estates for which we have full data coverage. We repeat this procedure 10,000 times and regress the random allocation of hollænderier on distance to Sofiendal and the full set of covariates to check whether the true distribution of hollænderier is better explained by distance to Sofiendal than a given random allocation. In figure 4 we plot the t-values of distance Sofiendal from these 10,000 regressions. The figure shows that Distance Sofiendal does indeed explain the true location of hollænderier significantly better relative to the vast majority of the 10,000 random allocations. We see this as a further piece of evidence that the hollænderier did in fact spread in the way we suggest from Sofiendal to the rest of the country.
24
Figure 4: Testing the relation between Distance Sofiendal and the distribution of hollænderier 0 3 -3 0,3 0,6 0,9 1,2 1,5 1,8 2,1 2,4 2,7 3,3 3,6 3,9 4,2 4,5 -4,5 -4,2 -3,9 -3,6 -3,3 -2,7 -2,4 -2,1 -1,8 -1,5 -1,2 -0,9 -0,6 -0,3 t-value (Distance Sofiendal)
Random Elite Elite 1782
Note: The red bars show frequencies of t-values of Distance Sofiendal from regressing (randomly allocated) 1782 on Distance Sofiendal and the full set of covariates. The random allocation of the 224 hollænderier to the 791 estates was repeated 10,000 times. The vertical blue line indicates the t-value from the true distribution of hollænderier (t-value= -2.25). 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒
Finally, we test the importance of regional fixed effects in our two stage least square estimates as they could potentially inflate the estimated effects. Hence, we re-estimate our main results in Table 4 but exclude regional level fixed effects in all specifications. Table 11 presents the results of this test. Panel A shows the results without fixed effects while Panel B repeats the results from Table 4 for convenience. We see that while there is a tendency for the fixed effects to inflate our estimates they are very stable as covariates are subsequently included and importantly, all results remain highly significant at conventional levels. Thus, we conclude that our instrumental variable results are not driven by the inclusion of fixed effects for regions.
<< Table 11 around here >>
25
6. Conclusion
How did Denmark ‘get to Denmark’? To the extent that the country developed through agricultural cooperation, the present work suggests a striking answer, and one which is not particularly compatible with the usual narrative of hard working peasants and a democratic countryside. Thus, we have demonstrated that the reason for the extremely rapid spread of cooperative creameries in Denmark between 1882 and 1890 can be attributed to the spread of innovations, starting with the introduction of the Holstein System, by large landowning elites from Schleswig and Holstein over the preceding century. We have described based on the contemporary literature how these innovations spread throughout the country, and trickled-down to farmers beyond the large estates. Moreover, we have demonstrated empirically that areas with more hollænderier developed greater cow densities, revealing the spread of dairying around the country, and that the initial wave of cooperation was in areas which had been so treated. We also show that the historical persistence of the elites still played a role in 1914 and that our results are stable to a number of robustness tests.
On a less optimistic note, however, it should also be remembered that the process as a whole took well over a century. The institutions, technology, schools, etc. did not appear overnight, or within the first decade of cooperation. Farmers would not have known that their comparative advantage lay in dairying in the 1880s, and they would not even have had the cow densities for this to be the case, if the hollænderier had never existed. This has implications for understanding the reason why the attempt to transfer Danish-style cooperatives to other countries, such as Ireland in the 1890s (see e.g. Henriksen et al. 2015) and Iceland around the turn of the twentieth century (Jónsson 2012), as well as to developing countries more recently, were relative failures. For more than a century elites were initiating a whole package of reforms which eventually allowed the cooperatives to emerge and prosper. There has been a tendency to see cooperation as the solution to agricultural poverty, but this work suggests that this must be in combination with other reforms, which, at least initially, elites are probably better placed to implement than the peasants themselves.
26
References
Acemoglu, D., S. Johnson and J. Robinson (2001). ‘The Colonial Origins of Comparative Development: An Empirical Investigation.’ American Economic Review 91, 1167-1199.
Akçomak, I.S., D. Webbink and B. der Weel (2016). ‘Why Did the Netherlands Develop So Early? The Legacy of the Brethren of the Common Life ’. Economic Journal 126 (593), 821–860.
Amsden, A.H., A. DiCaprio, J.A. Robinson (2012, eds.). The Role of Elites in Economic Development. Oxford: Oxford University Press.
Andersen, B.S. (1963). ‘Staldgård – eller hollænderi? Forsøg på status ved 1782’. Historie/Jyske Samlinger, Ny række, 6.
Andersen, T.B., P.S. Jensen, and C.S. Skovsgaard (2016). ‘The heavy plough and the agricultural revolution in medieval Europe’. Journal of Development Economics 118, 133–149.
Appel, A. and N. Bredkjær (1924-32). ‘Kvæget’ in K. Hansen (ed.), Det danske landbrugs historie, 3. bind: Husdyrbruget. Copenhagen: G.E.C. Gads Forlag.
Banerjee, A. and L. Iyer (2005). ‘History, Institutions, and Economic Performance: The Legacy of Colonial Land Tenure Systems in India’. American Economic Review 95:4, 1190-1213.
Baten, J. and D. Juif (2014). ‘A story of large landowners and math skills: Inequality and human capital formation in long-run development, 1820–2000’. Journal of Comparative Economics 42, 375-401.
Begtrup, G. (1803). Beskrivelse over Agerdyrkningens Tilstand i Danmark. Vol. Bind 1 og 2: Sjælland og Møen. Copenhagen: A. & S. Goldins Forlag.
Begtrup, G. (1806). Beskrivelse over Agerdyrkningens Tilstand i Danmark. Vol. Bind 3: Fyen, Langeland og Ærø. Copenhagen: A. & S. Goldins Forlag.
Begtrup, G. (1808). Beskrivelse over Agerdyrkningens Tilstand i Danmark. Vol. Bind 5 og 6: Nørre Jylland. Copenhagen: A. & S. Goldins Forla.
Becker, S.O. and L. Woessmann (2009). ‘Was Weber wrong? A human capital theory of protestant economic history’. Quarterly Journal of Economics 124:2, 531-596.
Bieleman, J. (1996). Five Centuries of Farming: A Short History of Dutch Agriculture, 1500-2000. Wageningen Academic Publishers.
Borup, E.J. and Nørgaard, F. (1939). Den Danske Folkehøjskole gennem hundrede aar. Odense: Skandinavisk Bogforlag.
Bjørn, C. (1977). ‘Fællesmejerierne – En fase i Dansk Mejeribrugs Udvikling 1860-1890’. Bol og By 2:1, pp. 65-80.
Bjørn, C. (1988). Det danske landbrugs historie III: 1810-1914. Odense: Landbohistorisk Selskab.
27
Boje, P. (2016). ‘How Denmark became rich. On the history of the Danish Innovation System.’ In Entreprenørskap: i næringsliv og politikk, edited by Knut Sogner, Einar Lie and Håvard Brede Aven. 29-56. Oslo: Novus forlag.
Boserup, E. (1992). 'En kommentar til Thorkild Kjærgaards disputats om Danmarks grønne revolution'. Fortid og Nutid 1, pp. 54-57.
Bustos, P., B. Caprettini and J. Ponticelli (2016). ‘Agricultural productivity and structural transformation. Evidence from Brazil’. American Economic Review 106(6), 1320-1365.
Cameron, A.C., and P.K. Trivedi (2009). Microeconometrics: Methods and Applications. New York: Cambridge University Press.
Chen, S. and J. K.-s. Kung (2016). ‘Of Maize and men? the effect of a New World crop on population and economic growth in China’, Journal of Economic Growth 21(1), 71-99. Christensen (Hørsholm), C. (1886). Agrarhistoriske Studier, I. Danske Landboforhold under Enevælden. Copenhagen: J.H. Schubothes Boghandel.
Cinnirella, F. and E. Hornung (2016). ‘Landownership concentration and the expansion of education’. Journal of Development Economics 121, 135–152.
Comin, D., W. Easterly, E. Gong (2010). ‘Was the wealth of nations determined in 1000 B.C.?’ American Economic Journal: Macroeconomics 2:3, 69-97.
Conley, T. G., C. B. Hansen, and P. E. Rossi (2012). ‘Plausibly Exogenous’. Review of Economics and Statistics 94(1), 260-272.
Cook, Justin C. (2014a). ‘The role of lactase persistency in precolonial development’. Journal of Economic Growth 19, 369-406.
Cook, Justin C. (2014b). ‘Potatoes, milk, and the Old World population boom’. Journal of Development Economics 110, 123-138.
Dalgas, C. (1826). Bidrag til Kundskab om de danske Provindsers nærværende tilstand i oekonomisk henseende, foranstaltet, efter kongelig Befaling af Landhuusholdningsselskabet, 1ste Stykke. Vejle Amt. Copenhagen: Det Kongelige Landhusholdningsselskab.
Dall, T., P.S. Jensen and A. Naz (2014). ‘New crops, local soils and urbanization: Clover, potatoes and the growth of Danish market towns, 1672-1901’. EHES Working Papers in Economic History 65.
Dell, M. (2010). ‘The persistent effects of Peru’s mining mita’. Econometrica 78:6, 1863-1903.
Dell, M. (2012). ‘Path dependence in development: Evidence from the Mexican Revolution’. Manuscript (Harvard University).
Dittmar, J. (2011). ‘Information technology and economic change: The impact of the printing press,’ Quarterly Journal of Economics 126, 1133-1172.
28
Drejer, A.A. (1925-33). ‘Mejeribruget’ in K. Hansen (ed.), Det danske landbrugs historie, 4. bind: Bygninger – Mejeri – Redskaber. Copenhagen: G.E.C. Gads forlag.
Drejer, A.A. (1962). Den historiske baggrund for nutidens mejeribrug og et halvt århundredes organisationsarbejde. Copenhagen: Særtryk af ’Dansk Mejeribrug’.
Engerman, S.L. and K.L. Sokoloff (2002). ‘Factor Endowments, Inequality, and Paths of Development Among New World Economics’. Economia 3:1 (NBER Working Paper 9259).
Falbe Hansen, V.A. (1889). Stavnsbaands-Løsningen og Landboreformerne set fra Nationaløkonomiens Standpunkt. Copenhagen: G.E.C. Gad.
Feldbæk, Ole (1988). ‘Kongen bød --- Enevælden og reformerne’. In Claus Bjørn (ed.), Landboreformerne. Forskning og forløb, Copenhagen: Landbohistorisk Selskab.
Fernández, Eva (2014). ‘Trust, religion, and cooperation in western agriculture, 1880-1930’. Economic History Review 67:3, 678-698.
Frandsen, K. (1983). ‘Vang og tægt. Studier over dyrkningssystemer og agrarstrukturer i Danmarks landsbyer 1682-83’. Esbjerg, Forlaget Bygd.
Frandsen, K. (2005). ‘Ager og eng’. In J. Erichsen and M.V. Pedersen (eds.). Herregården: Drift og landskab. Volume 3, Copenhagen: Dansk Bygningsarv, 25-100.
Fukuyama, F. (2011). The Origins of Political Order: From Prehuman Times to the French Revolution. London: Profile Books.
FAO/IIASA, (2002). GAEZ (Global Agro-Ecological Zones): http://webarchive.iiasa.ac.at/Research/LUC/GAEZ/index.htm.
Galor, O., O. Moav and D. Vollrath (2009). ‘Inequality in Land Ownership, the Emergence of Human Capital Promoting Institutions, and Great Divergence’. Review of Economic Studies 76:1, 143-179.
Oded Galor and Ömer Özak, 2016. “The Agricultural Origins of Time Preference,” American Economic Review, 2016, 106(10): 3064–3103.
Henriksen, I. (1993). ‘The transformation of Danish agriculture 1870–1914’. In K. G. Persson, ed., The economic development of Denmark and Norway since 1870. Aldershot: Edward Elgar, pp. 153–177.
Henriksen, I. (1999). ‘Avoiding lock-in: Cooperative creameries in Denmark, 1882–1903’. European Review of Economic History 3:1, 57-78.
Henriksen, I., M. Lampe, and P. Sharp (2011). ‘The role of technology and institutions for growth: Danish creameries in the late nineteenth century’. European Review of Economic History 15, 475-493.
Henriksen, I., M. Lampe, and P. Sharp (2012). ‘The strange birth of liberal Denmark: Danish trade protextion and the growth of the dairy industry since the mid-nineteenth century.’ Economic History Review 65(2), 770-788.
29
Henriksen, I., E. McLaughlin, and P. Sharp (2015). ‘Contracts and cooperation: The relative failure of the Irish dairy industry in the late nineteenth century reconsidered.’ European Review of Economic History 19(4), 412- 431.
Hansen, B.K. (2006). Familie- og arbejdsliv på landet, ca. 1870-1900. En undersøgelse af ægteskabsdannelse, familieforøgelse og arbejdsliv med særlig henblik på ændringerne i kvindernes arbejde. Auning: Landbohistorisk Selskab.
Harris, C. D. (1954). ’The market as a factor in the localization of industry in the United States.’ Annals of the Association of American Geographers 44, 315-348.
Hertel, H. (1920). Det Kgl. Danske Landhusholdningsselskabs historie I: Selskabets historie i tiden fra 1769-1868. Copenhagen: August Bangs Boghandel.
Hornung, E. (2014). ’Immigration and the Diffusion of Technology: The Huguenot Diaspora in Prussia.’ American Economic Review 104 (1), 84-122.
Iversen, P.K. (1992). ‘Hollænderier i Sønderjylland i 1600- og 1700-årene’. Sønderjyske Årbøger 1992, 73-111.
Jensen, S.P. (1998). ‘Kobbelbrug, kløver og kulturjord’. Bol og by 1998:1, 36-59.
Jia, Ruixue (2014). ‘Weather shocks, sweet potatoes and peasant revolts in historical China’. Economic Journal 124, 92-118.
Jones, P.M. (2016). Agricultural Enlightenment. Knowledge, Technology, and Nature, 1750-1840. Oxford University Press.
Jónsson, G. (2012). ‘The Impossible Dream: Transferring the Danish Agricultural Model to Iceland’. In Guirao, F., F.M.B Lynch, and S.M. Ramirez Pérez (eds.), Alan S. Milward and a Century of European Change. Abingdon: Routledge.
Khaustova, E. and P. Sharp (2015). ‘A Note on Danish Living Standards through Historical Wage Series, 1731- 1913’. Journal of European Economic History 44:3, 143-72.
Kjærgaard, T. (1994). The Danish revolution, 1500-1800: an ecohistorical interpretation. Cambridge: Cambridge University Press.
Koed, J. (1997). Danmarks Jernbaner I 150 år. Kunst & Kultur. Nordjyllands Bogtrykkeri A/S.
Kohl, J.G. (1846). Reisen in Dänemark und den Herzogthümern Schleswig und Holstein, 1. F.A. Brockhaus.
Laing, S. (1852). Observations on the social and political State of Denmark, and the duchies of Sleswick and Holstein in 1851: being the third series of the Notes of a traveller on the social and political state of the European people. Longman, Brown, Green, and Longmans.
Lampe, M. and P. Sharp (2014). ‘Greasing the wheels of rural transformation? Margarine and the competition for the British butter market’. Economic History Review 67:3, 769-792.
30
Lampe, M. and P. Sharp (2015a). ‘How the Danes Discovered Britain: The International Integration of the Danish Dairy Industry Before 1880’. European Review of Economic History 19(4), 432-453.
Lampe, M. and P. Sharp (2015b). ‘Just Add Milk: A Productivity Analysis of the Revolutionary Changes in Nineteenth Century Danish Dairying,’ Economic History Review 68(4), 1132-1153.
Lampe, M. and P. Sharp (forthcoming). The Creation of a "Land of Milk and Butter": How Elites Created the Modern Danish Dairy Industry. Book manuscript.
Linvald, A. (1905-08). ’Antvorskov og Vordingborg Krongodser 1768-1774’, Fra Arkiv og Museum 3, 234-289.
Linvald, A. (1912). "Hvem ejede Danmarks Jord omkring Midten af det 18. Aarhundrede?". Historisk Tidsskrift 4(8): 147-58.
McLaughlin, E. and P. Sharp (2015). ‘Competition between organisational forms in Danish and Irish dairying around the turn of the twentieth century’. University of St. Andrews Discussion papers in Environmental Economics 2015-16.
Mokyr, J. (2009). The Enlightened Economy: An Economic History of Britain 1700-1850, New Haven: Yale University Press.
Nunn, N. and N. Qian (2011). ‘The potato’s contribution to population and urbanization: evidence from a historical experiment’. Quarterly Journal of Economics 126:2, 593-650.
Ó Grada, C. (1977). ‘The Beginnings of the Irish Creamery System, 1880-1914’. Economic History Review 30, 284- 305.
Olsson, O. and D.A. Hibbs Jr. (2005). ‘Biogeography and long-run economic development’. European Economic Review 49:4, 909-938.
O'Rourke, K.H. (1997). ‘European Grain Invasion, 1870-1913.’ Journal of Economic History 57:4, 775-801.
O'Rourke, K.H. (2006). ‘Late Nineteenth-Century Denmark in an Irish Mirror: Land Tenure, Homogeneity and the Roots of Danish Success.’ In National Identity and the Varieties of Capitalism: The Danish Experience, edited by John L. Campbell, John A. Hall and Ove K. Pedersen. 159-96. Montreal, Kingston: McGill-Queen’s University Press.
O'Rourke, K.H. (2007). ‘Culture, Conflict and Cooperation: Irish Dairyng Before the Great War.’ Economic Journal 117, 1357-79.
Pedersen, J. (1999). Teknologisk udvikling i maskinindustrien: Burmeister & Wain 1875–1939. Lyngby.
Petersen, M.C. (1885). ‘Beretning om andelsmejeriernes første tid i Vestjylland 1884-1885’. Foreningen af Jyske Landboforeninger.
Porskrog Rasmussen, C., et. al. (1987, ed.). Det danske godssystem – udvikling og afvikling 1500-1919. Aarhus: Forlaget HISTORIA.
31
Porskrog Rasmussen, C. (2003). Rentegods og hovedgårdsdrift: Godsstruktur og godsøkonomi i hertugdømmet Slesvig 1524-1770. Del I: Fremstilling. Aabenraa: Institut for grænseregionsforskning.
Porskrog Rasmussen, C. (2010a). ‘Innovative Feudalism: The development of dairy farming and Koppelwirtschaft on manors in Schleswig-Holstein in the seventeenth and eighteenth centuries.’ Agricultural History Review 58:2, 172-190.
Porskrog Rasmussen, C. (2010b). ‘A.G. Moltke og det holstenske kobbelbrug.’ Landbohistorisk Tidsskrift 2010:2, 9-48.
Prange, Wolfgang (1971). Die Anfänge der großen Agrarreformen in Schleswig-Holstein bis um 1771. Neumünster: Karl Wachholtz (Quellen und Forschungen zur Geschichte Schleswig-Holsteins 60).
Putterman, L. (2008). ‘Agriculture, diffusion and development: Ripple effects of the Neolithic Revolution’. Economica 75, 729-748.
Roholt, P. (2012). ‘Herregårdene i landskabet og på nettet – mere end 700 herregårde lige ved hånden’. Herregårdshistorie 7, 161-166.
Schröder-Lembke, Gertrud (1978). Studien zur Agrargeschichte, Stuttgart: Gustav Fischer (Quellen und Forschungen zur Agrargeschichte, 31).
Skrubbeltrang, F. (1953). ‘Agricultural Development and Rural Reform in Denmark’, FAO Agricultural Studies 22.
Skrubbeltrang, F. (1978). Det danske Landbosamfund 1500-1800. Odense: Den danske historiske Forening.
Statistics Denmark (1911). Statistiske Meddelelser, Fjerde Række, Syv og Tredivte Bind, Første Hæfte. Folkemængden 1. Februar 1911 i Kongeriget Danmark. Statens Statistiske Bureau.
Squicciarini, M. and N. Voigtländer (2016). ’Knowledge Elites and Modernization: Evidence from Revolutionary France,’ Forthcoming in Quarterly Journal of Economics.
32
Table 1: Summary Statistics and Variable Descriptions Variable N Mean SD Min Max Source* Description Cooperative creamery exposure , A 38,370 8.22 2.00 3.17 14.05 Inverse distance weighted sum of cooperative creameries, 1890 (mp) 1890 Elites 1782 38,370 132.23 70.81 42.20 580.35 B Inverse distance weighted sum of hollænderi estate hartkorn, 1782 (mp) Distance Sofiendal 38,370 161.40 77.52 0.41 288.54 C Distance from grid cell to the estate Sofiendal (km) Distance first cooperative creamery 38,370 131.05 74.89 0.46 421.33 C Distance from grid cell to the first cooperative creamery in Hjedding (km) Estates (mp) 38,370 10.06 1.94 2.62 16.02 D Inverse distance weighted sum of estates, 1770 (mp) Demesne share, 1680s 38,370 0.12 0.28 0.00 1.00 E Share of grid cell area owned by an estate (with or without a hollænderi), 1680s Crown estates (mp) 38,370 2.25 0.86 0.71 6.37 F Inverse distance weighted sum of crown estates (min 10 years, 1600-1800) (mp) Late innovators (mp) 38,370 0.11 0.08 0.04 2.14 G Inverse distance weighted sum of ‘late innovator estates’ (mp) Folk high school (mp), 1890 38,370 0.75 0.24 0.26 3.55 H Inverse distance weighted sum of folk high schools, 1890 (mp) Butter production, 1662 38,370 1.18 4.89 0.00 73.21 I Butter payments in 1662 - barrels per km2 land in the grid cell Clover share, 1805 38,370 0.18 0.36 0.00 1.00 J Share of grid cell area cultivated with clover, 1805 Barley suitability 38,370 57.02 17.53 0.00 92.50 K Barley suitability from GAEZ, FAO (2002) Field-grass-system, 1682 38,370 0.55 0.50 0.00 1.00 I =1 if field-grass-system in 1682 Distance coast 38,370 9.45 9.46 0.00 48.67 E Distance from grid cell to the nearest coast (km) Distance Copenhagen 38,370 178.39 73.68 0.55 297.14 C Distance from grid cell to Copenhagen (km) Population density, 1787 38,010 21.58 96.17 0.00 7920.3 L Parish population density in 1787 Market town (mp) 38,370 0.90 0.20 0.47 3.15 C Inverse distance weighted sum of market towns (mp) Distance rail, 1890 38,370 10.37 20.15 0.00 177.31 M Distance from grid cell to the nearest rail road, 1890 (km) Distance Ox Road 38,370 64.81 67.49 0.00 368.30 C Distance from grid cell to the nearest Ox Road (km)
* Data sources: A) Own work, based on Bjørn (1988), B) own work, based on Andersen (1963), Christensen (1886), and Roholt (2012), C) own work, D) Christensen (1886), and Roholt (2012), E) own work, based on hiskis.dk, F) own work based on the Danish Center for Estate Research, G) own work based on Hertel (1920) – History of the Royal Danish Agricultural Society, the Royal Danish Agricultural Society , 1769-1868 H) own work based on Borup (1939) I) own work based on Frandsen (1983) J) own work based on Kjærgaard (1991), K) own work based on GAEZ, FAO (2002), L) Population count 1787, Statistics Denmark (1911) , M) own work based on Koed (1997)
33
Table 2: Main Results (OLS) – Cooperative Creamery Exposure 1890 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Elites 1782 0.01743*** 0.01710*** 0.01624*** 0.00639*** 0.00458*** 0.00469*** 0.00243*** (0.00064) (0.00112) (0.00112) (0.00095) (0.00095) (0.00094) (0.00086) [0.00316] [0.00260] [0.00272] [0.00295] [0.00260] [0.00262] [0.00243]
Distance first cooperative creamery -0.01589*** -0.01280*** -0.01324*** -0.01065*** -0.01525*** (0.00166) (0.00147) (0.00119) (0.00121) (0.00112)
Estates (mp) 0.43166*** 0.30008*** 0.30859*** 0.26660*** (0.02274) (0.02172) (0.02041) (0.01951)
Demesne share, 1680s -0.10556*** -0.09860*** -0.08036*** -0.04646** (0.02699) (0.02494) (0.02341) (0.02035)
Crown estates (mp) 0.45723*** 0.38210*** 0.22579*** (0.05187) (0.05022) (0.04701)
Late innovators (mp) 1.02982*** 0.72001*** 0.64265*** (0.28510) (0.24936) (0.22324)
Folk high school (mp), 1890 0.76982*** 0.67661*** 0.39155*** (0.09186) (0.09058) (0.08181)
Butter production, 1662 0.00083 0.00056 (0.00078) (0.00073)
Clover share, 1805 0.04172 0.01690 (0.03484) (0.03065)
Barley suitability 0.00293*** 0.00550*** (0.00097) (0.00087)
Field-grass-system, 1682 -0.35827*** -0.36490*** (0.06043) (0.05080)
Distance coast 0.02696*** 0.01487*** (0.00251) (0.00238)
Distance Copenhagen -0.01688*** (0.00113)
Population density, 1787 -0.00040*** (0.00007)
Market town 0.04290 (0.07866)
Distance rail, 1890 -0.01782*** (0.00236)
Distance Ox Road -0.01200*** (0.00132)
Constant 5.91323*** 3.33979*** 9.82393*** 7.82893*** 7.42727*** 6.16450*** 18.25437*** (0.08849) (0.06521) (0.69005) (0.62739) (0.51385) (0.53036) (0.88956)
FE (Region) No Yes Yes Yes Yes Yes Yes N 38370 38370 38370 38370 38370 38370 38010 Adj R2 0.382 0.805 0.822 0.857 0.878 0.888 0.905 Parish level clustered standard errors in parentheses, Conley standard errors correcting for spatial autocorrelation within 50 km in squared brackets, * p < 0.1, ** p < 0.05, *** p < 0.01 34
Table 3: Main Results (IV, First stage) – Elites 1782 and Distance to Sofiendal (1) (2) (3) (4) (5) (6) (7) Dependent variable: Elites 1782 Distance Sofiendal -0.6575*** -0.6473*** -0.6659*** -0.4206*** -0.4071*** -0.4719*** -0.7571*** (0.0155) (0.0366) (0.0383) (0.0295) (0.0353) (0.0332) (0.1161)
Distance first cooperative creamery -0.1969*** -0.0513 -0.0444 -0.1530*** -0.0998*** (0.0419) (0.0324) (0.0323) (0.0320) (0.0384)
Estates (mp) 12.2277*** 12.1003*** 12.3435*** 12.2350*** (0.6724) (0.7739) (0.7327) (0.7791)
Demesne share 3.1931*** 3.2440*** 2.8024*** 3.1049*** (0.7527) (0.7460) (0.7468) (0.7391)
Crown estates (mp) 1.3574 -0.7378 -1.7443 (1.7865) (1.7960) (1.8319)
Late innovators (mp) -23.7641*** -20.6686*** -20.2987*** (7.6383) (7.5099) (7.5279)
Folk high school (mp), 1890 1.9137 0.7805 2.5192 (1.9618) (1.8326) (1.7879)
Butter production, 1662 0.0155 0.0133 (0.0243) (0.0242)
Clover share, 1805 0.2781 -0.0939 (0.9960) (1.0087)
Barley suitability 0.1256*** 0.0882*** (0.0252) (0.0262)
Field-grass-system, 1682 -9.5859*** -9.2537*** (1.4132) (1.4260)
Distance coast -0.4205*** -0.3408*** (0.0552) (0.0613)
Distance Copenhagen 0.3300*** (0.1171)
Population density, 1787 -0.0044* (0.0023)
Market town -0.1469 (2.4761)
Distance rail, 1890 -0.0601 (0.0787)
Distance Ox Road 0.0928** (0.0385)
Constant 238.3455*** 166.4430*** 249.7980*** 110.2120*** 103.5156*** 153.3186*** 111.5902*** (3.5016) (6.9252) (17.6343) (13.7017) (15.1500) (15.5743) (29.1330)
FE (Region) No Yes Yes Yes Yes Yes Yes N 38370 38370 38370 38370 38370 38370 38010 Parish level clustered standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
35
Table 4: Main Results (IV, Second stage) – Cooperative Creamery Exposure 1890 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Elites 1782 0.0196*** 0.0504*** 0.0520*** 0.0634*** 0.0555*** 0.0434*** 0.0489*** (0.0008) (0.0030) (0.0028) (0.0051) (0.0053) (0.0037) (0.0077) [0.00450] [0.00942] [0.00835] [0.01448] [0.01376] [0.00951] [0.01813]
Distance first cooperative creamery -0.0110*** -0.0132*** -0.0137*** -0.0083*** -0.0046* (0.0020) (0.0022) (0.0019) (0.0017) (0.0026)
Estates (mp) -0.3791*** -0.3340*** -0.1891*** -0.3401*** (0.0831) (0.0760) (0.0564) (0.1024)
Demesne share, 1680s -0.2382*** -0.2218*** -0.1553*** -0.1687*** (0.0572) (0.0513) (0.0414) (0.0507)
Crown estates (mp) 0.0769 0.1569* 0.1858* (0.1096) (0.0889) (0.0973)
Late innovators (mp) 2.1067*** 1.5142*** 1.7044*** (0.5448) (0.4295) (0.5047)
Folk high school (mp), 1890 0.3981*** 0.4330*** 0.3662*** (0.1011) (0.0837) (0.0887)
Butter production, 1662 -0.0003 -0.0005 (0.0013) (0.0014)
Clover share, 1805 -0.0084 -0.0135 (0.0538) (0.0584)
Barley suitability -0.0013 -0.0002 (0.0013) (0.0017)
Field-grass-system, 1682 0.1044 0.1382 (0.0867) (0.1165)
Distance coast 0.0337*** 0.0331*** (0.0032) (0.0053)
Distance Copenhagen 0.0011 (0.0036)
Population density, 1787 0.0000 (0.0001)
Market town -0.0459 (0.1430)
Distance rail, 1890 -0.0154*** (0.0042)
Distance Ox Road -0.0147*** (0.0023)
Constant 5.6289*** 1.8447*** 6.2341*** 7.7170*** 7.6630*** 5.5294*** 11.7066*** (0.1053) (0.1411) (0.8812) (0.8946) (0.7859) (0.7025) (2.0592)
FE (Region) No Yes Yes Yes Yes Yes Yes N 38370 38370 38370 38370 38370 38370 38010 First stage F 1795.176 312.670 301.853 203.510 133.026 202.601 42.517 Parish level clustered standard errors in parentheses, Conley standard errors correcting for spatial autocorrelation within 50 km in squared brackets, * p < 0.1, ** p < 0.05, *** p < 0.01
36
Table 5: Main Results (IV, Second stage) – with Standardized Coefficients (“Beta Coefficients”) (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Elites 1782 .69400*** 1.78643*** 1.84322*** 2.24516*** 1.96617*** 1.53697*** 1.73159***
Distance first cooperative -.41233*** -.49572*** -.51324*** -.31204*** -.17426* creamery
Estates (mp) -.36723*** -.32353*** -.18324*** -.32897***
Demesne share,1680s -.033161*** -.03088*** -.02163*** -.02347***
Crown estates (mp) 0.03304 0.06740* .07995*
Late innovators (mp) .08193*** .05889*** .06652***
Folk high school (mp), 1890 .09509*** .10342*** .08673***
Butter production, 1662 -.00083 -.00113
Clover share, 1805 -.00150 -.00241
Barley suitability -.01160 -.00178
Field-grass-system, 1682 .02601 .03442
Distance coast .15969*** .15612***
Distance Copenhagen .03990
Population density, 1787 .00081
Market town .00464
Distance rail, 1890 -.15589***
Distance Ox Road -.49698***
37
Table 6: Channel (IV, Second stage) – Cow Density 1837 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cow Density, 1837 Elites 1782 0.0921*** 0.0867*** 0.0942*** 0.1119*** 0.0689*** 0.0672*** 0.1015*** (0.0040) (0.0083) (0.0090) (0.0134) (0.0146) (0.0154) (0.0353)
Distance first cooperative creamery 0.0320*** 0.0365*** 0.0265*** 0.0145** 0.0296** (0.0060) (0.0069) (0.0063) (0.0065) (0.0142)
Estates, 1700 (mp) -0.6554*** -0.7166*** -0.5747*** -1.4651*** (0.2319) (0.2054) (0.2180) (0.4486)
Demesne share, 1680s -0.7471 0.1459 -0.1400 0.4266 (1.1278) (0.9920) (0.9570) (1.1934)
Crown estates (mp) 1.8101*** 0.6153 0.7042 (0.5386) (0.5070) (0.5671)
Late innovators (mp) 8.8885*** 6.6957*** 7.8451** (2.6674) (2.5780) (3.4008)
Folk high school, 1890 1.1100 1.1186 0.6832 (0.8437) (0.8003) (0.9633)
Butter production, 1662 0.3354*** 0.3493** (0.1270) (0.1421)
Clover share, 1805 -0.9530 -0.8459 (0.7885) (0.8983)
Barley suitability 0.0596*** 0.0622*** (0.0110) (0.0138)
Field-grass-system, 1682 -3.3601*** -2.5389** (0.8389) (1.0082)
Distance coast -0.1898*** -0.2223*** (0.0275) (0.0324)
Distance Copenhagen -0.0013 (0.0141)
Population density, 1787 -0.0011 (0.0022)
Market town -5.3662*** (1.4821)
Distance rail road -0.1032*** (0.0343)
Distance Ox Road -0.0659*** (0.0155)
Constant 3.8554*** 11.2005*** -2.1116 -2.9002 1.3891 3.0318 37.7817*** (0.6344) (2.1281) (3.4101) (3.6707) (3.2781) (3.3874) (9.5380)
FE (Region) No Yes Yes Yes Yes Yes Yes Parishes (N) 1641 1641 1641 1641 1641 1641 1641 First stage F 906.398 341.133 306.667 164.838 103.898 84.817 20.103 Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
38
Table 7: Robustness (IV, Second stage) – Cooperative Creamery Exposure 1914 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1914 Elites 1782 0.0064*** 0.0662*** 0.0695*** 0.0787*** 0.0731*** 0.0562*** 0.0544*** (0.0015) (0.0049) (0.0045) (0.0080) (0.0083) (0.0057) (0.0080)
Distance first cooperative creamery -0.0230*** -0.0248*** -0.0253*** -0.0165*** -0.0102*** (0.0029) (0.0030) (0.0028) (0.0025) (0.0029)
Estates (mp) -0.3097** -0.2743** -0.0894 -0.2335** (0.1276) (0.1228) (0.0893) (0.1106)
Demesne share, 1680s -0.3217*** -0.3111*** -0.2190*** -0.2156*** (0.0759) (0.0716) (0.0568) (0.0587)
Crown estates (mp) -0.1047 0.0556 0.0846 (0.1465) (0.1148) (0.1075)
Late innovators (mp) 2.8477*** 2.0542*** 2.2468*** (0.7730) (0.6004) (0.6231)
Folk high school (mp), 1890 0.4413*** 0.5332*** 0.2905*** (0.1449) (0.1182) (0.1052)
Butter production, 1662 0.0002 0.0002 (0.0017) (0.0015)
Clover share, 1805 0.0215 0.0040 (0.0730) (0.0670)
Barley suitability -0.0019 0.0040** (0.0019) (0.0020)
Field-grass-system, 1682 0.5637*** 0.4409*** (0.1366) (0.1376)
Distance coast 0.0482*** 0.0397*** (0.0043) (0.0059)
Distance Copenhagen -0.0047 (0.0038)
Population density, 1787 0.0000 (0.0001)
Market town 0.4416*** (0.1688)
Distance rail, 1890 -0.0225*** (0.0054)
Distance Ox Road -0.0386*** (0.0027)
Constant 14.9747*** 3.4341*** 12.6026*** 13.8216*** 13.8093*** 10.2632*** 25.5625*** (0.2134) (0.2244) (1.2816) (1.2318) (1.1492) (1.0190) (2.3287)
FE (Region) No Yes Yes Yes Yes Yes Yes N 38370 38370 38370 38370 38370 38370 38010 First stage F 1795.176 312.670 301.853 203.510 133.026 202.601 42.517 Parish level clustered standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
39
Table 8: Robustness (IV, Second stage) – with Standardized Coefficients (“beta coefficients”) (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1914 Elites 1782 0.1521*** 1.5667*** 1.6458*** 1.8636*** 1.7305*** 1.33052*** 1.2904***
Distance first cooperative -0.5754*** -0.6200*** -0.6342*** -0.4133*** -0.2552*** creamery
Estates (mp) -0.2005** -0.1775** -0.0579 -0.1515**
Demesne share, 1680s -0.0299*** -0.0289*** -0.0204*** -0.0201***
Crown estates (mp) -0.0301 0.0160 0.0244
Late innovators (mp) 0.0740*** 0.0534*** 0.0588***
Folk high school (mp), 1890 0.0740*** 0.0851*** 0.0461***
Butter production, 1662 0.0003 0.0004
Clover share, 1805 0.0259 0.0005
Barley suitability -0.0110 0.0235**
Field-grass-system, 1682 0.0938*** 0.0736***
Distance coast 0.1526*** 0.1256***
Distance Copenhagen -0.1167
Population density, 1787 0.0004
Market town 0.0299***
Distance rail, 1890 -0.1521***
Distance Ox Road -0.8748***
40
Table 9: Parish Level Results (IV, Second stage) – Cooperative Creamery Exposure 1890 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Elites 1782 0.020*** 0.053*** 0.052*** 0.048*** 0.039*** 0.032*** 0.045*** (0.001) (0.003) (0.003) (0.004) (0.004) (0.004) (0.009)
Distance first cooperative creamery -0.004* -0.004** -0.006*** -0.006*** -0.001 (0.002) (0.002) (0.002) (0.002) (0.004)
Estates, 1770 (mp) 0.122* 0.100* 0.165*** -0.201* (0.067) (0.057) (0.053) (0.116)
Demesne share, 1680s -1.926*** -1.749*** -1.385*** -1.122*** (0.332) (0.282) (0.238) (0.344)
Crown estates (mp) 0.378** 0.278** 0.266 (0.154) (0.126) (0.166)
Late innovators (mp) 2.212*** 1.461** 2.237** (0.768) (0.647) (0.978)
Folk high school, 1890 -0.250 -0.224 -0.379 (0.243) (0.201) (0.275)
Butter production, 1662 0.037 0.066 (0.032) (0.042)
Clover share, 1805 -0.344* -0.454* (0.194) (0.259)
Barley suitability -0.001 0.004 (0.003) (0.004)
Field-grass-system, 1682 -0.856*** -0.501* (0.210) (0.290)
Distance coast 0.051*** 0.037*** (0.007) (0.009)
Distance Copenhagen -0.004 (0.004)
Population density, 1787 0.000 (0.000)
Market town -0.164 (0.329)
Distance rail road -0.005 (0.010)
Distance Ox Road -0.033*** (0.004)
Constant 5.676*** 1.815*** 3.304*** 3.429*** 4.377*** 4.529*** 15.575*** (0.140) (0.672) (1.117) (1.051) (0.911) (0.830) (2.797)
FE (Region) No Yes Yes Yes Yes Yes Yes Parishes (N) 1675 1675 1675 1675 1675 1675 1675 First stage F 932.554 348.808 312.922 170.219 108.669 88.848 25.552 Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
41
Table 10: Estate Level Results (IV, Second stage) – Cooperative Creamery Exposure 1890 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Elite 1782 8.13652*** 19.93444*** 19.45107*** 20.12016*** 19.78238*** 12.24757*** 11.52997** (0.92323) (4.70804) (5.00953) (5.31063) (5.15796) (3.41786) (5.08719)
Distance first cooperative creamery -0.00404 -0.00302 -0.00244 -0.00489 -0.00604 (0.00744) (0.00780) (0.00777) (0.00511) (0.00657)
Estate size -0.01377 -0.01296 -0.00469 -0.00346 (0.00881) (0.00859) (0.00537) (0.00492)
Crown estate 0.54374 0.34869 0.34582 (0.66120) (0.43188) (0.43778)
Late innovator 3.22361 2.06005 2.09942 (2.87729) (1.82270) (1.78802)
Folk high school, 1890 -1.72729 -1.18905 -1.12460 (1.13602) (0.72532) (0.77846)
Butter production, 1662 0.23548* 0.22610* (0.13606) (0.12765)
Pasture suitability 0.00026 0.00001 (0.00021) (0.00020)
Demesne clover share, 1805 -0.17802 -0.33733 (0.61182) (0.59273)
Barley suitability -0.00010 0.00008 (0.00014) (0.00013)
Field-grass-system, 1682 -1.19250 -1.48014* (0.78098) (0.78708)
Distance coast 0.11958*** 0.07888*** (0.02661) (0.02917)
Distance Copenhagen -0.01155 (0.00810)
Population density, 1787 -0.00380 (0.00803)
Distance Market town -0.01540 (0.03085)
Distance rail road, 1890 -0.00498 (0.02893)
Distance Ox Road -0.04698*** (0.01153)
Constant 6.27175*** -3.69486 -2.82643 -2.83392 -2.64082 0.67929 6.85949 (0.28624) (3.48531) (4.27626) (4.41280) (4.30145) (2.96982) (4.74731)
FE (Region) No Yes Yes Yes Yes Yes Yes Estates (N) 810 810 810 810 810 796 791 First stage F 88.009 18.391 15.525 14.711 15.031 13.194 5.079 Standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
42
Table 11: Main Results [T4] (IV, 2. stage) Without Fixed Effects – Cooperative Creamery Exposure 1890 and Elites 1782 (1) (2) (3) (4) (5) (6) (7) Dependent variable: Cooperative creamery exposure, 1890 Panel A
Elites 1782 0.0196*** 0.0196*** 0.0273*** 0.0245*** 0.0159*** 0.0158*** 0.0106*** (0.0008) (0.0008) (0.0007) (0.0007) (0.0007) (0.0009) (0.0009)
FE (Region) No No No No No No No First stage F 1795.176 1795.176 2799.355 2555.892 970.787 823.029 840.414
Panel B Elites 1782 0.0196*** 0.0504*** 0.0520*** 0.0634*** 0.0555*** 0.0434*** 0.0489*** (0.0008) (0.0030) (0.0028) (0.0051) (0.0053) (0.0037) (0.0077)
FE (Region) No Yes Yes Yes Yes Yes Yes First stage F 1795.176 312.670 301.853 203.510 133.026 202.601 42.517
Full set of covariates (T4, Col 7) Yes Yes Yes Yes Yes Yes Yes N 38370 38370 38370 38370 38370 38370 38010 Parish level clustered standard errors in parentheses, * p < 0.1, ** p < 0.05, *** p < 0.01
43
Appendix A – Estate Level Descriptives
Table A1: Summary Statistics and Variable Descriptions Variable N Mean SD Min Max Description Cooperative creamery exposure, 810 8.52 2.06 4.50 14.00 Inverse distance weighted sum of cooperative creameries, 1890 (mp) 1890 Elites 1782 810 0.28 0.45 0.00 1.00 =1 if the estate had a hollænderi Distance Sofiendal 810 147.48 82.67 0.00 285.80 Distance to the estate of Sofiendal (km) Distance first cooperative creamery 810 141.69 58.65 3.57 262.84 Distance to the first cooperative creamery in Hjedding (km) Estate size 810 44.68 35.22 6.00 325.61 Historical measure of the value of the estate in 1770 (Domain size and soil suitability) Crown estate 810 0.23 0.42 0 1 =1 if crown ownership at least 10 years during 1600-1800 Late innovator 810 0.01 0.09 0.00 1.00 =1 if the estate was one of the ‘late innovators’ Folk high school, 1890 810 0.07 0.26 0.00 1.00 =1 if folk high school in the parish before 1890 Butter production, 1662 810 1.40 1.45 0.00 7.94 Butter payments in 1662 - barrels per km2 land in the parish of the estate Pasture suitability 810 63.17 12.70 26.80 77.70 Pasture suitability at the estate or the nearest geographical coverage of FAO Demesne clover share, 1805 796 0.35 0.43 0.00 1.00 Share of grid cell area cultivated with clover, 1805 Barley suitability 810 60.04 19.56 15.00 92.50 Barley suitability at the estate or the nearest geographical coverage of FAO Field-grass-system, 1682 810 0.46 0.48 0.00 1.00 Share of parish using field-grass-system in 1682 Distance coast 810 7.17 7.04 0.00 40.67 Distance to the nearest coast (km) Distance Copenhagen 810 168.22 73.00 8.32 290.79 Distance to Copenhagen (km) Population density, 1787 804 21.87 21.40 0.00 416.90 Parish level population density in 1787 Market town 810 11.36 5.66 0.03 37.14 Distance to the nearest market town (km) Distance rail, 1890 810 6.85 6.23 0.01 37.59 Distance to the nearest rail road, 1890 (km) Distance Ox Road 810 67.85 56.66 0.00 202.64 Distance to the Ox Road (km)
44
Appendix B – Parish Level Descriptives
Table B1: Summary Statistics and Variable Descriptions Variable N Mean SD Min Max Description Cooperative creamery exposure, 1847 8.42 2.17 2.67 13.86 Inverse distance weighted sum of cooperative creameries, 1890 (mp) 1890 Elites 1782 1847 147.14 88.31 42.07 1356.06 Inverse distance weighted sum of hollænderi estate hartkorn, 1782 (mp) Cow density, 1837 1673 17.68 9.10 0.00 103.30 Number of cows per km2, 1837 Distance Sofiendal 1847 148.64 78.04 2.80 287.88 Distance from parish center* to the estate Sofiendal (km) Distance first cooperative creamery 1847 138.20 67.28 3.01 420.23 Distance from parish center to the first cooperative creamery in Hjedding (km) Estates (mp) , 1770 1847 10.20 2.05 2.63 14.73 Inverse distance weighted sum of estates, 1770 (mp) Demesne share, 1680s 1847 0.12 0.19 0.00 0.99 Share of parish owned by an estate (with or without a hollænderi), 1680s Crown estates (mp) 1847 2.37 0.88 0.71 5.33 Inverse distance weighted sum of crown estates (min 10 years, 1600-1800) (mp) Late innovators (mp) 1847 0.12 0.08 0.04 1.14 Inverse distance weighted sum of ‘late innovator estates’ (mp) Folk high school (mp), 1890 1847 0.05 0.23 0.00 1.00 =1 if folk high school in parish before 1890 Butter production, 1662 1847 1.26 1.62 0.00 13.58 Butter payments in 1662 - barrels per km2 land in the parish Clover share, 1805 1847 0.22 0.35 0.00 1.00 Share of parish cultivated with clover, 1805 Barley suitability 1808 56.77 20.85 0.00 92.50 Average parish barley suitability, GAEZ (FAO 2002) Field-grass-system, 1682 1754 0.45 0.48 0.00 1.00 Share of parish with field-grass-system in 1682 Distance coast 1847 7.73 7.35 0.00 44.59 Distance from parish center to the nearest coast (km) Distance Copenhagen 1847 168.40 74.04 1.79 295.12 Distance from parish center to Copenhagen (km) Population density, 1787 1708 35.91 227.00 0.00 7920.25 Parish population density in 1787 Market town 1722 0.04 0.20 0.00 1.00 =1 if market town in parish Distance rail, 1890 1847 8.44 18.61 0.01 176.80 Distance from parish center to the nearest rail road, 1890 (km) Distance Ox Road 1847 69.67 66.33 0.01 367.28 Distance from parish center to the Ox Road (km) *: The parish center is defined as the location of the church. Some rural parishes located near a market town were served by the church in the market town and hence had no church. Hence, for such parishes with no church prior to the introduction of the first hollænderier in the 1760s we use the geographical center (centroid). For parishes with more than one church older than 1760 we use the location of oldest church as the parish center.
45
Appendix C – Plausibly exogenous We test the exogeneity of our instrumental variable using the ‘plausibly exogenous’ framework of Conley et al. (2012). Instead of claiming complete exogeneity of instruments the framework tests the degree to which the instrument is endogenous and tests the 2SLS results given the potential level of endogeneity. In our case we test the degree to which our IV Distance Sofiendal directly explains the geographical distribution of cooperative creameries. If this degree is low we can arguable claim that our results are robust to the potential endogeneity.
We use the ‘Union of Confidence Intervals’ (UCI) approach and assume that the direct impact of our IV is somewhere between zero and the upper 95% confidence interval value from the reduced form estimate with the full set of controls ( [ 0.0433; 0]).
𝛿𝛿 ∈ − Figure C1: Testing the plausible exogeneity
For the distance to Sofiendal to directly explain away our results delta would have to be -0.032 or below as indicated in the figure above (blue line). As the reduced form estimate of distance Sofiendal is -0.037 (green line) the direct channel should be 86% of the entire reduced form estimate. This seems highly unlikely and hence we trust our results to be robust to the potential level of endogeneity of our instrument.
46